R v Secretary of State for Work and Pensions

The various appellants in each of two appeals, which have been heard together, challenge the lawfulness of provisions relating to what is known as the revised benefit cap.

The original benefit cap was introduced by section 96(1) of the Welfare Reform Act 2012 (“the 2012 Act”). Pursuant to it, the Housing Benefit Regulations 2006, SI 2006/213, (“the 2006 Regulations”) were amended so as to provide, in regulation 75A, that, if a household’s total entitlement to specified welfare benefits were otherwise to exceed an annual limit, its entitlement should be capped at that limit. The original cap came into force on 15 April 2013.

In R (SG) v Secretary of State for Work and Pensions [2015] UKSC 16, [2015] 1 WLR 1449, this court, by a majority of three to two, dismissed an appeal by three lone mothers and three of their children against a decision that provisions relating to the original cap did not discriminate against women in the enjoyment of their possession of welfare benefits and so were not unlawful. I will refer to the SG case as the first benefit cap case.

In its manifesto for the general election which took place on 7 May 2015 the Conservative Party proposed that any Conservative government would introduce legislation for a revised benefit cap which would cap specified benefits at a lower level. Following the party’s victory in that election the government introduced, and Parliament enacted, the Welfare Reform and Work Act 2016 (“the 2016 Act”). By making amendments to the 2012 Act, the 2016 Act introduced the revised cap, which came into force on 7 November 2016. In making provision for the original cap, the earlier version of the 2012 Act had, in section 96(5) to (7), provided for the annual limit, at which the welfare benefits were to be capped, to be specified in regulations and to be determined by reference to the estimated average net earnings of a working household in Britain; and the amended 2006 Regulations had specified that, for couples and lone parents, the annual limit was £26,000, being a figure apparently determined in that way.

But the amendments wrought by the 2016 Act have replaced those provisions; and, for the purposes of the revised cap, they identify the annual limits in the 2012 Act itself, namely in a new section 96(5A). The effect of the subsection, when read with a new regulation 75CA inserted into the 2006 Regulations by regulation 2(3) of the Benefit Cap (Housing Benefit and Universal Credit) (Amendment) Regulations 2016 (SI 2016/909) (“the 2016 Regulations”) is that, for couples and lone parents, the annual limits are reduced to £23,000 if they reside in Greater London and to £20,000 if they reside elsewhere. How were these reduced figures calculated? Clearly the yardstick of average net earnings of a working household was abandoned - otherwise the figures would not have come down. The government’s Impact Assessment dated August 2016 relating to the 2016 Act (“the IA”) suggested that the reduced figures were calculated by reference to the fact that 40% of households earn less than them. But, say the appellants, the only arguably relevant figures would relate to the total income of those households, inclusive in particular of benefits. The amendments made in 2016 provide no automatic adjustment of the limits for inflation; and the reduced figures have already lost 5% of their real value. But a new section 96A of the 2012 Act requires the Secretary of State to review them at least once during each Parliament. The welfare benefits subject to the cap, which prior to the amendments to the 2012 Act were left to be specified in regulations, are also now specified in the Act itself, namely in section 96(10). Among others, the benefits there specified include child benefit, child tax credit, housing benefit and income support.

Various features of the scheme which applied to the original cap have been retained for application to the revised cap. By regulation 75D of the 2006 Regulations, it is for the local authority to implement the cap by reducing payment of housing benefit accordingly. By regulation 75F, those in receipt of certain benefits (now including, pursuant to amendment by the 2016 Regulations, a carer’s allowance and a guardian’s allowance) are exempt from the cap even if they also receive benefits which are specified in section 96(10) as being subject to it. And, most importantly, by regulation 75E(2), those entitled to working tax credit are exempt from the cap. Under regulation 4(1) of the Working Tax Credit (Entitlement and Maximum Rate) Regulations 2002 (SI 2002/2005) a single person (which here includes a lone parent) is entitled to working tax credit if, among other things, she or he undertakes work for at least 16 hours each week. A couple, on the other hand, is entitled to it if, among other things, they undertake work for at least 24 hours each week, provided that one of them does so for at least 16 of those hours.

In the IA the government stated that its introduction of the revised cap had three aims:

to improve the fairness of the social security system and to increase public confidence in its fairness, particularly in relation to the government’s objective not to reward a non-working family with an income in the form of welfare benefits which exceeded that of an average working family;

to make fiscal savings which would enable the government to redirect its limited resources for better deployment elsewhere; and

to incentivise the parents or parent in a non-working family to obtain work on the basis in particular that an ethic of work within a family inculcated better outcomes for its children.

The IA identified the incentivisation of work as the main aim. The ability of parents to escape the cap by undertaking work for not less than the specified number of hours is described by the government as a key exemption and is therefore central to the design of the scheme.

The basic argument on behalf of the appellants is that, in introducing the revised cap, the government, through Parliament, has unlawfully discriminated both against the lone parents of young children, whose ability to work is severely curtailed by their childcare obligations, and against the young children of lone parents.

In the DA case there are five appellants. Three of them are lone parent mothers. They all care for children of various ages. At the outset of the proceedings the children of two of the mothers included a child aged under two. Those two children then aged under two are the other appellants. Those two mothers had each suffered a reduction in benefits as a result of the revised cap. At the outset of the proceedings the third mother was pregnant.

In the DS case there are eleven appellants. Two of them are lone parent mothers. They both care for children of various ages. The first mother cares for five children, each of whom is an appellant. The second cares for four children, each of whom is also an appellant. At the outset of the proceedings none of these nine children was aged under two but three of them were aged under five. Both of the mothers had suffered a reduction in benefits as a result of the revised cap.

Both sets of appellants primarily cast their claim of unlawful discrimination under the Human Rights Act 1998 (“the 1998 Act”). Their secondary challenge to the scheme for the revised cap is that, in its application to them, it is irrational at common law; but, if the primary claim fails, the application of the scheme to them will not be irrational so the secondary challenge will not further be addressed.

In the DA case the appellant mothers contend

In the DA case the appellant children contend

In the DS case the appellant mothers contend

that their entitlement to welfare benefits falls within the ambit of their rights both under article 1 p 1 and under article 8;

that, in that they have the status either of lone parents or, as a fall-back, of lone parents of children aged under five, they have the right under article 14 to claim that their rights under article 1 p 1 and article 8 have not been secured without discrimination;

that under the scheme they are subject to such different treatment in comparison with other adults, particularly with dual-care parents or with lone parents all of whose children are aged at least five, that, unless the different treatment of them is justified, the law requires them to have the same treatment; and

that the government has failed to justify their subjection to such different treatment and that therefore it has unlawfully discriminated against them, whether directly or indirectly.

In the DS case the appellant children contend

that they have an interest in the welfare benefits to which their mothers are entitled and that it falls within the ambit of their own rights under article 8 or, if not, under article 1 p 1;

that, in that they have the status of children of lone parents or, in the case of three of them and as a fall-back, that they have the status of children aged under five of lone parents, they have the right under article 14 to claim that their rights under article 8 or, if not, under article 1 p 1 have not been secured without discrimination;

that under the scheme they are subject to such different treatment in comparison with other children, particularly children of dual-care parents or children aged at least five of lone parents, that, unless the different treatment of them is justified, the law requires them to have the same treatment; and

that, in particular in the light of an alleged breach on its part of article 3 of the UNCRC, the government has failed to justify their subjection to such different treatment and that therefore it has unlawfully discriminated against them, whether directly or indirectly.

In response to the above contentions the government concedes only that the entitlement of the two sets of appellant mothers to welfare benefits falls within the ambit of their rights under article 1 p 1. It disputes every other contention.

Unlike the DS case, the DA case has been the subject of adjudication on the merits in the lower courts. By an order dated 22 June 2017, [2017] EWHC 1446 (Admin), [2017] PTSR 1266, Collins J upheld the claims of the DA claimants by declaring that the 2006 Regulations, as amended by the 2016 Regulations, unlawfully discriminated against lone parents of children aged under two and against children aged under two of lone parents. But, by an order dated 15 March 2018, [2018] EWCA Civ 504, [2018] PTSR 1606, the Court of Appeal (Sir Patrick Elias who gave the main judgment and Sir Brian Leveson, President of the Queen’s Bench Division, who gave a short concurring judgment; McCombe LJ dissenting) set aside the order of Collins J. In effect the court dismissed the claims and granted permission to appeal to the Supreme Court.

On 26 March 2018, thus 11 days after the order of the Court of Appeal in the DA case, Lang J heard the DS case. Mr Drabble QC, on behalf of the DS claimants, submitted to her that the dismissal of their claims was not strictly mandated by the Court of Appeal’s decision in the DA case but he conceded that it placed significant hurdles in their way. In light of the fact that the decision in the DA case was to be reviewed in the Supreme Court, he persuaded the judge to dismiss their claims without inquiry into their merits and to grant a leap-frog certificate under section 12 of the Administration of Justice Act 1969 to the effect that an application on their part to the Supreme Court for leave to appeal directly to it would be justified. In due course such an application was made to this court and granted. The consequence is that there has been no lower court review of the evidence filed in the DS case.

Although the law of discrimination is inherently difficult, it is impossible to avoid the conclusion that, for various reasons, the court’s examination of the issues raised in these appeals has been unnecessarily cumbersome and complicated.

The judgments of each of the five members of the court in the first benefit cap case, all of alleged relevance, proceed in all for 269 paragraphs and their combined effect has been a matter of acute and protracted debate in this court and elsewhere.

The three judgments of the Court of Appeal in the DA case, now before this court, proceed for 184 paragraphs.

The written cases presented to this court in the two appeals by the three principal parties and the three interveners, all of high legal quality, proceed across 357 pages.

The evidence filed in both appeals proceeds across more than 3,000 pages.

119 authorities are presented to the court for consideration.

The oral argument has continued for two and a half days.

In the above circumstances the compilation of this judgment has had to be surgical. Reference to all the arguably relevant evidence and submissions would have submerged it. As it is, I am disappointed with myself in having failed to contain it within fewer than 91 paragraphs.

The impact of the revised cap has been broadly as follows:

As at August 2017, the benefits of 68,000 households had been reduced by reference to the revised cap.

52,000 (ie 77%) of those households would not have suffered reduction by reference to the original cap; so the revision of the cap has had a substantial effect.

Of the 68,000 households which suffered the revised cap, 49,000 (ie 72%) were lone parent households.

Since 90% of lone parents are women, 44,000 (ie 65%) of the households which suffered the revised cap were lone female parent households.

Of the 68,000 households which suffered the revised cap, 37,000 (ie 54%) were lone parent households with a child aged under five and, of those, 17,000 (ie 25% of the total) were lone parent households with a child aged under two. The cap has therefore had a major impact on the former group, of which the latter are a significant subgroup.

Families with multiple children, thus in receipt of higher amounts of child benefit and child tax credits, are more likely to be capped. As at February 2018, 74% of capped households (not here differentiated between dual-care and lone parent households) had at least three children.

Has the revised cap incentivised those on benefits to work? The government accepts that the statistical evidence is sparse; and it is inappropriate to address it in detail. It suffices to say:

In putting forward its expectations for the revised cap in the IA, the government suggested that 41% of those potentially subject to it would be more likely to work in order to escape it than those not potentially subject to it.

But the statistic turns out to mean that the number of those more likely to work in order to escape the cap is 41% larger than the already small group, namely only 11% of all capped households, who would have moved into work in any event. Translated into numbers, it means that only about one capped household out of 20 such households (ie 5%) was considered likely to move into work in order to escape it. In relation, however, to that one capped household out of 20, the appeals require the court to consider whether it was more likely to be a dual-care household than a lone parent household, in particular a lone parent household with a child aged under five or indeed aged under two.

So, in relation to incentivisation, the government relies less on statistics and more on what are said to be the obvious financial advantages of working. These advantages are scarcely in dispute. Evidence on behalf of the appellants in the DS case suggests, by way of example in relation to one of the mothers, that, when capped, her annual household benefits were £20,000 but that, were she to have worked for 16 hours each week earning £17,000 net, her net annual income would have risen to £32,000 because her benefits would have been reduced by only £5,000.

Irrespective, however, of the financial advantages for a parent who works hours sufficient to claim working tax credit and thus to escape the cap, how practicable is it for a lone parent, in particular a lone parent of a child aged under five or indeed aged under two, to do so?

(a) Is it reasonable to divert the lone parent from caring for such

Central to the government’s response to these questions is its provision, on certain conditions, of free childcare for 30 hours per week during term-time under the Childcare (Early Years Provision Free of Charge) (Extended Entitlement) Regulations 2016 (S1 2016/1257). The IA stressed its availability and estimated it to be worth about £5,000 pa per child. The trouble is that the provision extends to free care only for three and four year olds and also, albeit limited to 15 hours per week, for certain two year olds in families in receipt of specified benefits. This, no doubt, explains why in the DA case the appellants are members of families which include a child aged under two. Mr Wise QC on their behalf therefore points to the grave difficulty which confronts lone parents in that group in accessing care so that they can work. To this the government responds that the financial advantages of escaping the cap by work are so substantial, as explained above, that these lone parents, if in work, can afford to pay for childcare out of their overall income inclusive of benefits and that they are substantially assisted in doing so in the computation of their working tax credit.

But Mr Wise draws a wider point from the limited extent of the provision for free childcare: it betokens (he says) a considered governmental conclusion that it is not in the interests of lone parents of children aged under two, nor in particular in the interests of those children, that their parents should be diverted from caring for them.

Here Mr Wise and Mr Drabble join in making an allied point. It relates to the conditions attached to the receipt of income support, which is likely to be a major constituent of the welfare benefits paid to a lone parent. The aim of the conditions is to make it easier for her to find work when (but only when) her youngest child has attained the age of five. One condition relates to the period when she has a child aged one or two and it requires her to attend work-focussed interviews about every six months. Another relates to the period when she has a child aged three or four and it requires her to engage in some training or other work-related activity in preparation for future work. The sanction for failing to comply with a condition is a reduction in income support. Once all her children have attained the age of five, in other words are of school age, the lone parent not in work must claim jobseeker’s allowance instead of income support and, to that end, must demonstrate that she is available to do a limited amount of work and that she is actively seeking it. The point made on behalf of all the appellants is that at the heart of the carefully calibrated regime of attaching conditions to the receipt by lone parents of income support is a recognition by the government that it is wrong to expect them actually to work until all their children have attained school age; and that to cap their benefits for failure to work before all their children have attained school age flies in the face of that policy decision.

The government’s defence of its application of the revised cap to lone parents in the circumstances identified in these appeals relates in significant part to the provision for the possible making to them of a Discretionary Housing Payment (“a DHP”). Provision for DHPs is made in section 69 of the Child Support, Pensions and Social Security Act 2000 and in regulations made under it. A power to make a DHP is conferred on local authorities and, as the title implies, it must relate to housing costs. So, when a cap requires a local authority to reduce housing benefit below, or further below, the level of the recipient’s rent, there is the facility for it to make a DHP to cover the balance. Central government provides local authorities both with an annual fund out of which to make DHPs and with a guidance manual in relation to their distribution of them. A broad discretion is conferred upon the local authorities. There is no appeal against a refusal to make a DHP but there is, with whatever degree of difficulty, an opportunity to challenge it by way of judicial review. DHPs are intended to cover many more situations of hardship than those created by the cap, including in particular hardship created by the provisions addressed by this court in R (MA) v Secretary of State for Work and Pensions [2016] UKSC 58, [2016] 1 WLR 4550 (“the bedroom tax case”); and it is clear that they are mainly intended to alleviate temporary hardship and, for those subject to the cap, are intended, in the words of the IA, to manage “the transition for various customers whilst they make the necessary changes to adapt to the application of the benefit cap”. The manual however now includes within a list of possible recipients a household which “contains … a child … under two years of age where childcare is a barrier to getting work”.

In the light in particular of the provision within the scheme for the exercise of discretion on the part of local authorities in the making of DHPs, how far should this court rely on them for alleviation of the worst effects of the cap on lone parents within the appellant groups? The government refers powerfully to the fact that five out of the six appellant mothers in these appeals have at one stage been in receipt of DHPs. The problem is that, as the government accepts, there is limited data about the extent to which capped households are rescued by DHPs. Both the appellants in the DA case and Shelter, as an Intervener in the appeals, present evidence of divergence in

the degree of complexity, sometimes bewildering, with which local authorities surround the making of an application for a DHP;

the time which they take to process it;

the period for which, subject only to some further award, they agree to make a DHP in order to alleviate a cap, awards of indefinite duration being unknown and most being subject to a maximum of 12 weeks; and

the extent to which any award of a DHP is large enough to cover the shortfall in housing benefit imposed by the cap.

Of the other two aims of government in introducing the revised cap, that of making fiscal savings has scarcely been pressed. The IA forecast that in the year just past (2018-19) the revised cap would save the state £110m. But the figure did not include the operational cost of implementing the cap nor the cost of the support, in particular through DHPs, otherwise provided to capped claimants, all now estimated at £68m this year. In context the net figure appears to be 0.03% of the state’s overall annual expenditure on welfare benefits for those of working age.

Does the revised cap inflict poverty on those subject to it? The answer is hotly contested. But since the government cannot sensibly argue that the computation of welfare benefits is intended to provide a family with more than it needs, it follows that a reduction of those benefits will provide it with less than it needs. Of course the concept of needs is to some extent elastic: they can be assessed with somewhat greater or lesser stringency. But the government does not seek to argue that the lower figures set for the revised cap have been reached by reference to any scale of needs. Equally, in a speech in 2016 relied on by the government, Mr Cameron, then the Prime Minister, acknowledged that the effect of welfare benefits was “to push people’s incomes just above the poverty line”. It follows that a substantial reduction in them pulls their incomes well down below the poverty line. In my view there are sound reasons for accepting the evidence given by the Child Poverty Action Group in the DS case that the effect of the cap is to reduce a family well below the poverty line, judged by the generally accepted measure of less than 60% of median UK income equivalent to the size of the household.

There ensues striking evidence adduced on behalf of the DA appellants about the effect on children of an early life of poverty. Professor Atkinson, the former Children’s Commissioner for England, echoing evidence from Jonathan Bradshaw, Professor of Social Policy at York University, offers this summary:

In M v Secretary of State for Work and Pensions [2006] UKHL 11, [2006] 2 AC 91, Lord Nicholls observed in para 14 that

In Petrovic v Austria (2001) 33 EHRR 14 the European Court of Human Rights (“the ECtHR”) held that a refusal to pay a father, as opposed to a mother, a parental leave allowance fell within the ambit of his rights under article 8 because, as explained in para 27, the allowance enabled a parent to stay at home to look after the children and so affected the way in which family life was organised. In Okpisz v Germany (2006) 42 EHRR 32 it held that a decision no longer to pay child benefits to certain aliens fell within the same ambit. In the bedroom tax case, cited at para 30 above, this court held at para 49 and unanimously, that the so-called bedroom tax, by which housing benefit was capped by reference to rules about the number of bedrooms which a family needed, fell within the same ambit. Earlier, in the first benefit cap case, the court had no need to consider whether the original cap fell within the ambit of article 8. Nevertheless the government relies on doubts about it which Lord Reed expressed in para 79. The difficulty is that, as the Intervener, Just Fair, suggests, Lord Reed there seems to have equated the ambit of article 8 with interference with rights under it, which, with respect, may not be the usual analysis.

In the DA case Collins J and the Court of Appeal both held that the revised cap fell within the ambit of the rights under article 8 of the claimant mothers and children. I have no doubt that they were correct - and of course the same applies to the claimants in the DS case. The effect of the provisions for the cap may be that the mother goes to work and escapes it; if so, her children below school age have to be cared for in some other way. Or the effect may be that the cap is imposed, with a variety of possible results: that, as expressly suggested by the government to be an option, the family, no doubt with great difficulty, has to move to cheaper accommodation; or that the mother builds up rent arrears and so risks eviction or otherwise falls into debt; or that, like one of the DA mothers, she has to cease buying meat for the children; or, as in cases recorded by Shelter, that she has to go without food herself in order to feed the children or has to turn off the heating. Whatever their individual effect, provisions for a reduction of benefits to well below the poverty line will strike at family life.

The government argues, if faintly, that in the DA case the Court of Appeal was wrong to conclude that the claimants, in other words both the lone parents and the children, had a “status” on the ground of which they might seek to complain under article 14 of discrimination in the enjoyment of their substantive Convention rights. The government submits, for example, that the parents are women, who admittedly enjoy a status under article 14, and that it is therefore inappropriate for them to seek to shoehorn themselves into some other status. The submission is difficult to understand: it is of the essence of the parents’ case in the DA appeal and of what I regard as a significant part of their case in the DS appeal that they are lone parents of children aged under two or under five, and that the discrimination lies in the difference between their situation and that of others subject to the cap. The government proceeds to submit that the situation of the appellants can be transitory in that a parent may not be a lone parent for ever and that a child will not remain aged under two (or under five) for long. But there is no ground for concluding that a status for the purpose of article 14 has to be permanent. Some of the examples of status given in article 14 itself can change - religion, political opinion, even sex.

In Mathieson v Secretary of State for Work and Pensions [2015] UKSC 47, [2015] 1 WLR 3250, this court referred in para 21 to previous authority that the concept of status generally comprised personal characteristics and that inquiry into it should concentrate “on what somebody is, rather than what he is doing or what is being done to him”; it observed in para 22 that, if the complaint of discrimination fell within the ambit of a Convention right, the ECtHR was reluctant to conclude that the complainant had no relevant status; and it held in para 23 that, as a “severely disabled child in need of lengthy in-patient hospital treatment”, the appellant’s deceased son had had a status within the meaning of article 14. In R (Stott) v Secretary of State for Justice [2018] UKSC 59, [2018] 3 WLR 1831, this court recently conducted a detailed examination of the meaning of “other status” in article 14. In the event all members of the court other than Lord Carnwath confirmed that its meaning was broad; and they proceeded to hold that a prisoner subject to a particular type of sentence thereby had the status which under article 14 enabled him to allege that its effect had been to discriminate against him in the enjoyment of his rights under article 5 of the Convention. The present appellants assert statuses more obviously composed of personal characteristics than were those recognised in the cases of Mathieson and Stott; and I have no doubt that all of them have the requisite status in the terms set out in paras 13(b), 14(b), 15(b) and 16(b) above.

In R (A) v Secretary of State for Health [2017] UKSC 41, [2017] 1 WLR 2492, I noted in para 29 that the claimants complained that they should have been treated in the same way as a specified group but in para 30 that they had turned their argument inside out in complaining alternatively that they should have been treated in a different way from another specified group. I added in para 31 that in that case the alternative presentation added only an extra level of unwelcome complexity. Nevertheless the concept of discrimination is, as Sir Patrick said in para 17 of his judgment in the DA case, underpinned by the fundamental principle not only that like cases should be treated alike but also that different cases should be treated differently. And in some cases, unlike the A case but exemplified by that in the ECtHR of Thlimmenos v Greece (2000) 31 EHRR 12, the natural formulation of the complaint is indeed that the complainants have been treated similarly to those whose situation is relevantly different, with the result that they should have been treated differently.

I have sought to describe in paras 13(c), 14(c), 15(c) and 16(c) above the way in which the various appellants before the court formulate their complaints of discrimination.

The DA appellants primarily complain that, in applying the revised cap (a) to lone parents of children aged under two and (b) to children aged under two of lone parents (together, “the DA cohorts”), the government has treated them similarly to others to whom it has applied the cap but whose situation is relevantly different from theirs. So the DA appellants say that, unless the similar treatment can be justified, the government should have treated them differently by exempting them from the cap.

But, like the appellants in the A case, the DA appellants can also turn their complaint inside out. They can point to the exemption from the cap granted to those in receipt of a carer’s allowance (paid to those who for at least 35 hours a week care for an adult on specified benefits) and of a guardian’s allowance (paid to those who bring up a child of deceased parents). So in the alternative the DA appellants can complain that, in applying the cap to themselves, the government has treated them differently from carers and guardians to whom it has not applied the cap but whose situation is relevantly similar to theirs. So, the DA appellants can say that, unless the different treatment can be justified, the government should have treated them similarly by exempting them from it.

Although the alternative formulation of the complaint of the DA appellants has arguable merit, I have no doubt that the natural way of analysing their complaint accords with their primary formulation of it: it is of discrimination of the type explained in the Thlimmenos case, namely that, by subjecting them to the revised cap, the government has treated the DA cohorts similarly to a specified group whose situation is relevantly different from theirs and thus that, subject to justification, it should have treated them differently from it.

I confess that I have found it less easy to understand the way in which the DS appellants formulate their complaint. They contend that the revised cap represents discrimination, both direct and indirect, which violates the Convention rights of all lone parents (and/or women because 90% of lone parents are women) and of all children of lone parents. A group of all lone parents would of course include lone parents with children all aged between five and 18, ie all of school age; and so too a group of all children of lone parents would include children of school age. But the evidence of the DS appellants has scarcely been directed to the effect of the cap on households with children all of school age. Mr Drabble is no doubt entitled, by reference to the statistics set out in para 22 above, to complain of the particular effect of the cap on all lone parents and thus on women; but, insofar as they are lone parents of children all of school age, it is already obvious that the government can justify it. In my view the complaint of the DS appellants which the court should proceed to address is their fall-back complaint, namely that the cap violates the Convention rights (a) of all lone parents with a child aged under five and (b) of all children aged under five of lone parents (together, “the DS cohorts”). Although, for reasons unclear, the DS appellants formulate their fall-back complaint only reluctantly in accordance with the Thlimmenos case, such seems to me to be, as in the DA case, its natural formulation, namely that, by subjecting them to the revised cap, the government has treated the DS cohorts similarly to a specified group whose situation is relevantly different from theirs and thus that, subject to justification, it should have treated them differently from it.

The question then arises: what is the specified group which the government is said to have treated similarly to the DA and the DS cohorts? As here, the identification of a comparator group can be difficult. In the present case is the proper comparator

dual-care parents with a child aged under two or under five; or

lone parents without a child aged under two or under five; or

all others subjected to the revised cap?

Blessed is simplicity. The complaint made by the appellants is that their cohorts should not have been subjected to the revised cap. The natural corollary is, as Mr Wise contends, that they are comparing their cohorts with all others subjected to the cap: so the natural comparator is the group at (c). Nevertheless, in arguing that there has been an objectionable similarity of treatment between the DA and the DS cohorts, on the one hand, and all others subjected to the cap, on the other, the appellants may seek to highlight their objection by reference to subgroups, such as those at (a) and (b), whose situations are alleged to be relevantly different.

In DH v Czech Republic (2008) 47 EHRR 3 the Grand Chamber of the ECtHR said in para 175 that “discrimination means treating differently, without an objective and reasonable justification, persons in relevantly similar situations”. Re-cast to cover the type of discrimination recognised in the Thlimmenos case, the proposition is that it means treating similarly, without an objective and reasonable justification, persons in relevantly different situations. In Carson v United Kingdom (2010) 51 EHRR 13 the Grand Chamber explained in para 61 what was meant by the absence of objective and reasonable justification: “in other words, if it does not pursue a legitimate aim or if there is not a reasonable relationship of proportionality between the means employed and the aim sought to be realised”.

Clarity of language aids clarity of thought. It is worthwhile to stress, as the court did in the Mathieson case in para 24, that the frequent reference to “justified discrimination” in the domestic discussion of the concept is, as a matter of law, the expression of a contradiction in terms. As the terminology long favoured by the Grand Chamber shows, justification negatives the very existence of discrimination.

In the DH case the Grand Chamber proceeded to explain in para 177 that, once the applicant had shown a difference in treatment of persons in relevantly similar situations, the burden of proof lay on the state to establish that it was justified; and in para 178 that what shifted the burden on to the state was “prima facie evidence”.

There is clear prima facie evidence that - in the terms of the re-cast proposition - the DA and the DS cohorts are in a relevantly different situation from those others who have been treated similarly to them by their common subjection to the revised cap. For it appears

that, in the case of a lone parent of a child below school age, in particular of a child below the age of two, it is contrary to the interests both of herself, of her child and of the family as a whole that she should in effect be constrained to work also outside the home;

that, by the conditions which it has attached to the receipt of income support, the government has itself decided that it is contrary to their interests;

that, irrespective of whether it is contrary to their interests for her to be so constrained, the extra difficulty, beyond that faced by others subjected to the cap, which confronts such parents in finding not only suitable work but also suitable childcare is plain;

that, in the case of a child aged under two, the absence of any free childcare further increases that difficulty;

that the incidence of failure of those represented by the DA and the DS cohorts to escape the cap, namely in the case of the wider DS cohort 54%, and in the case of the narrower DA cohort 25%, of all those who suffer it, demonstrates its disproportionate impact upon them; and

that, while the effect of the cap on all households who suffer it is to reduce their income below the poverty line, poverty has a disproportionate effect on the young children within these cohorts, stunting major aspects of their development in the long term as well as in the short term.

In the first benefit cap case Lord Reed said in para 14:

May I suggest, with respect, that Lord Reed may there have identified the focus of the justification too widely? He described it as “the legislation governing the cap”. In A v Secretary of State for the Home Department [2004] UKHL 56, [2005] 2 AC 68, Lord Bingham of Cornhill stated in para 68:

I conclude that what the government has to justify in the present case is its failure to amend the 2006 Regulations so as to provide for exemption of the DA and DS cohorts from the revised cap. The Secretary of State does not appear to challenge this conclusion.

This court has been proceeding down two different paths in its search for the proper test by which to assess the justification under article 14 for an economic measure introduced by the democratically empowered arms of the state. In retrospect this duality has been unhelpful. I regret having contributed to it.

The considerations which have informed the mapping of both paths is best explained by two citations. First, from the judgment of Lord Hope of Craighead in In re G (Adoption: Unmarried Couple) [2008] UKHL 38, [2009] AC 173, para 48:

Lord Reed then completed para 93 by adding

The appropriateness of an inquiry into whether the adverse effects of certain measures are manifestly without reasonable foundation is firmly rooted in the jurisprudence of the ECtHR. In James v United Kingdom (1986) 8 EHRR 123, in which it rejected the challenge to the legislation in England and Wales for leasehold enfranchisement, that court, in plenary session, held at para 46 that it should respect the judgment of the national legislature as to what was in the public interest unless it was manifestly without reasonable foundation. And in Stec v United Kingdom (2006) 43 EHRR 47, para 52, which it repeated word for word in Carson v United Kingdom (2010) 51 EHRR 13, para 61, the Grand Chamber, addressing complaints of discrimination arising out of the rules for entitlement to social security benefits, held that it should respect the national legislature’s determination of where the public interest lay when devising economic or social measures unless it was manifestly without reasonable foundation. It explained that this more benign approach to the establishment of justification for the adverse effects of a rule flowed from the margin of appreciation which was wide in this area of decision-making.

I now accept that the weight of authority in our court mandates inquiry into the justification of the adverse effects of rules for entitlement to welfare benefits by reference to whether they are manifestly without reasonable foundation.

In Humphreys v Revenue and Customs Comrs [2012] UKSC 18, [2012] 1 WLR 1545, the court rejected a complaint that a rule for entitlement to child tax credit discriminated against men on the basis that the different treatment of men which resulted from the rule was not manifestly without reasonable foundation. In her judgment, with which the other members of the court agreed, Lady Hale said in para 19 that in the context of state benefits the normally strict test for justification of the effect of a rule alleged to be discriminatory on grounds of sex “gives way”; but she added in para 22 that it did not follow that such a rule should escape careful scrutiny.

The possible mapping of a different path emerged in the judgment of Lord Mance in In re Recovery of Medical Costs for Asbestos Diseases (Wales) Bill [2015] UKSC 3, [2015] AC 1016. But the subject-matter was not the entitlement to welfare benefits. It was a proposed bill in the National Assembly of Wales for costs incurred by NHS Wales to be cast upon identified groups. The court’s decision was that the bill fell outside the assembly’s legislative competence. But Lord Mance proceeded to consider, in passing, whether the bill would have infringed the rights of the groups under article 1 p 1. This required him to consider how the court should assess whether a deprivation of property was justified. In this regard he referred in para 45 to the four stages of a conventional inquiry into justification. In para 52 he held that the first three stages (which require the establishment of a legitimate aim of the measure, of a rational connection of the measure to the aim and of an inability to achieve it less intrusively) could be addressed by whether the contentions in support of the measure were manifestly without reasonable foundation; but that the fourth stage (which requires the establishment of a fair balance between all the interests in play) fell for decision by the court, although it might pay significant respect to the balance favoured by those responsible for the measure.

A month after delivery of its judgments in the Wales case the court delivered its judgments in the first benefit cap case. It proceeded on the agreed basis that, if the analysis reached the stage at which the effect of the impugned provisions fell to be justified, the appellants had to establish that it was manifestly without reasonable foundation; and, by a majority, the court held that they had failed to satisfy this agreed criterion. In their dissenting judgments Lady Hale and Lord Kerr duly applied the agreed criterion, albeit with opposite results. But in para 209 of her judgment Lady Hale referred to para 52 of Lord Mance’s judgment in the Wales case and observed that a benefit cap closely resembled a deprivation of property; and in para 210 she floated the idea that, in the absence of agreement upon the criterion, it might have been possible to limit its application to the stages of the conventional inquiry referable to the aim of the provisions and to exclude its application to the final stage referable to its fair balance and overall proportionality.

Almost two years later the court delivered its judgments in the bedroom tax case, cited in para 30 above. Two of the three conjoined appeals concerned claims that the effect of rules for the computation of housing benefit was to discriminate against disabled people in the enjoyment of their rights under article 8 and/or article 1 p 1. Giving the main judgment, Lord Toulson recorded in para 28 the primary contention of the claimants in the first appeal as having been that the Court of Appeal had erred in asking whether the treatment of which they complained was manifestly without reasonable foundation. In paras 29 to 38 he then at length set out reasons in support of his conclusion, in which all the other members of the court concurred, that the Court of Appeal had not erred when, in assessing justification for the effect of the rules on the claimants, it had asked itself that single question.

Several months after delivery of the judgments in the bedroom tax case, the court delivered its judgments in the A case cited in para 40 above. The case concerned not welfare benefits but the government’s refusal, partly on grounds of cost, to exercise its power to require the NHS in England to provide free abortion services to women usually resident in Northern Ireland. One of the arguments on behalf of the women was that the effect of its refusal was to discriminate against them in the enjoyment of their rights under article 8 of the Convention. I gave a judgment, with which Lord Reed and Lord Hughes agreed, in which I rejected the argument. Lady Hale and Lord Kerr gave judgments in which they upheld it. It was in the course of my judgment, in para 33, that I cited the judgment of Lord Mance in the Wales case and asserted it to have become clear that, of the four aspects of an inquiry into justification under the Convention of the effect of a measure of economic or social policy, the fourth, relating to a fair balance, fell to be answered by the court for itself and not by reference to whether it was manifestly without reasonable foundation.

We may put aside consideration of whether the government decision impugned in the A case was of a character, unlike its rules of entitlement to welfare benefits, which made my suggested approach to its justification sound in law. For, even if so, I expressed myself too widely. Even though none of the other members of the court, including those in dissent, took issue with what I said, I take sole responsibility for it. Probably also emboldened by Lady Hale’s observations in the first benefit cap case, I reached too quickly for the observations of Lord Mance in the Wales case. For by then there was - and there still remains - clear authority both in the Humphreys case and in the bedroom tax case for the proposition that, at any rate in relation to the government’s need to justify what would otherwise be a discriminatory effect of a rule governing entitlement to welfare benefits, the sole question is whether it is manifestly without reasonable foundation. Let there be no future doubt about it.

How does the criterion of whether the adverse treatment was manifestly without reasonable foundation fit together with the burden on the state to establish justification, explained in para 50 above? For the phraseology of the criterion demonstrates that it is something for the complainant, rather than for the state, to establish. The rationalisation has to be that, when the state puts forward its reasons for having countenanced the adverse treatment, it establishes justification for it unless the complainant demonstrates that it was manifestly without reasonable foundation. But reference in this context to any burden, in particular to a burden of proof, is more theoretical than real. The court will proactively examine whether the foundation is reasonable; and it is fanciful to contemplate its concluding that, although the state had failed to persuade the court that it was reasonable, the claim failed because the complainant had failed to persuade the court that it was manifestly unreasonable.

Article 3 of the UNCRC provides:

What does the concept of the best interests of the child in article 3.1 encompass? In the Mathieson case, at para 39, this court approved a suggestion which Lord Carnwath had made in para 105 of the first benefit cap case to the effect that authoritative guidance was to be found in para 6 of General Comment No 14 (2013) of the UN Committee on the Rights of the Child. There the committee had suggested that the concept had three dimensions:

a substantive right of the child to have his or her best interests assessed as a primary consideration when different interests are being considered in order to reach a decision on the issue at stake;

an interpretative principle, irrelevant to the present appeals; and importantly;

a rule of procedure that, whenever a decision is to be made that will affect an identified group of children, the decision-making process must include an evaluation of the possible impact of the decision on them.

In the light in particular of the Mathieson case, the government cannot deny that the committee’s analysis is authoritative guidance in relation to the dimensions of the concept in article 3.1. It can submit only, and correctly, that the guidance is not binding even on the international plane and that, while it may influence, it should, as mere guidance, never drive a conclusion that the article has been breached.

The UNCRC also provides:

under article 26(1) that “States Parties shall recognize for every child the right to benefit from social security … and shall take the necessary measures to achieve the full realization of this right in accordance with their national law”;

The ECtHR has made it clear that, where relevant, the content of another international convention, in particular one relating to human rights such as the UNCRC, should inform interpretation of the Convention: Neulinger v Switzerland (2010) 54 EHRR 1087, paras 131 and 132.

It follows that, when relevant, the content of the UNCRC can inform inquiry into the alleged violation of article 14 of the Convention, when taken with one of its substantive rights.

But in what circumstances is any breach of article 3.1 of the UNCRC relevant to an alleged violation of article 14? The question was addressed by each of the five members of this court in the first benefit cap case, in which the suggested violation of article 14 lay in the cap’s alleged discrimination against women in the enjoyment of their right to possession of welfare benefits under article 1 p 1. The answers were as follows:

Lord Reed assumed, rather than held, in para 88 that the cap breached article 3.1 of the UNCRC but held at para 89 that such breach was irrelevant to the alleged discrimination against women.

Lord Carnwath held in paras 122 to 128 that the cap did breach the article but held in paras 125 to 131 that such breach was irrelevant to the alleged discrimination.

Lord Hughes held in para 146 that any such breach was irrelevant to the alleged discrimination and in paras 148 to 155 that in any event the cap did not breach the article.

Lady Hale held in para 224 that any breach of the article was relevant to the alleged discrimination and in paras 226 to 229 that the cap did breach it.

Lord Kerr, like Lady Hale, held in para 262 that the cap did breach the article and in paras 263 to 268 that the breach was relevant to the alleged discrimination.

In the present case the complaint of discrimination differs from the complaint in the first benefit cap case. The adult victims of the alleged discrimination are now cast not merely as women but as lone parents of children below school age. Moreover these children are now cast as further victims of it in their own right. And, although the lone parents repeat their complaint of discrimination in the enjoyment of their rights under article 1 p 1 of the Convention, both they and their children now complain of it in relation to the enjoyment of their respective rights to respect for their family life under article 8.

In explaining in the first benefit cap case that a breach, if any, of article 3.1 was irrelevant to the alleged discrimination, Lord Reed, Lord Carnwath and Lord Hughes each stressed in the paragraphs cited above that in their view the alleged discrimination could not be said to be directed against children. It is clear that the government cannot import their reasoning into the present proceedings. Equally it undertakes a mammoth task in maintaining the argument that, in setting the terms of the revised cap, it was not taking an action “concerning children” within the meaning of article 3.1. If valid in relation to the revised cap, the argument would have been valid in relation to the original cap. But it was rejected by Lord Carnwath, Lady Hale and Lord Kerr; and it was specifically upheld neither by Lord Reed nor by Lord Hughes. In para 107 Lord Carnwath referred further to General Comment No 14, namely to para 19 in which the committee explained that the duty under article 3.1 applies to all decisions on the part of public authorities which directly or indirectly affect children.

Insofar as in the present appeals the children themselves claim a violation of rights of their own under article 14, taken with article 8, their rights should be construed in the light of the UNCRC as an international convention which identifies the level of consideration which should have been given to their interests before subjecting their households to the revised cap.

But can the lone parents themselves also claim that their own rights under article 14, taken with article 8, must be construed in the light of the provision in the UNCRC for consideration of their children’s interests? The interests of the lone parents in play in the present appeals are indistinguishable from the interests of their children below school age. Their claim is as parents: so, without their children, it would not exist. Indeed their claim is as lone parents: so responsibility for their children in effect rests solely upon them. And their claim is to defend furtherance of their family life from the effects of a cap on benefits specifically computed by reference to the needs of their children and themselves taken together. Never more apt than to the present appeals is the observation of Lady Hale in Beoku-Betts v Secretary of State for the Home Department [2008] UKHL 39, [2009] AC 115, in para 4 that:

The claims of all the appellant cohorts under article 14, taken with article 8, therefore require the court to proceed to assess whether, in setting the terms of the revised cap, the government breached article 3.1 of the UNCRC. Were the court to hold that it had done so, what would the effect of it be? The overarching inquiry is whether its decision not to exempt the appellant cohorts from the cap was manifestly without reasonable foundation. As McCombe LJ observed, albeit more forcefully, in para 178 of his dissenting judgment in the DA case, a foundation for the decision not made in substantial compliance with article 3.1 might well be manifestly unreasonable.

In deciding upon the terms of the revised cap, did the government have regard, as a primary consideration, to the best interests of children below school age of lone parents and did it evaluate the possible impact of its decision upon them?

In answering this question within its overarching inquiry into the alleged violation of Convention rights, the court can, without constitutional impropriety, have regard to Parliamentary materials which explain the background to the government’s decision and in particular its policy objectives: Wilson v First County Trust Ltd (No 2) [2003] UKHL 40, [2004] 1 AC 816, paras 61 to 66.

It is worthwhile to preface an attempted answer to the question by adverting to two features of the Parliamentary discussion of the Bill which introduced the original cap, both briefly noted by Lord Reed in his judgment in the first benefit cap case at paras 29 and 40.

In May 2011 Ms Buck MP tabled an amendment before the Public Bill Committee of the House of Commons to the effect that households should be exempt from the cap if childcare costs outweighed earnings. She referred in particular to lone parents with four or five children, of whom one was aged under five. The government opposed the amendment, which Ms Buck withdrew.

In November 2011 the Lord Bishop of Ripon tabled an amendment before the Lords’ Grand Committee that lone parents of children aged under five should be exempt from the cap. He said:

On 8 September 2015, following presentation to Parliament of the Bill which included provision for the revised cap, the government published its Memorandum to the Joint Committee on Human Rights, in which, pursuant to its duty under section 19 of the 1998 Act, it stated that in its view the provision was compatible with Convention rights. In the memorandum the government

accepted in para 18 that the provision fell within the ambit of article 1 p 1 and arguably of article 8;

noted in para 19 the decision of this court in the first benefit cap case;

accepted in para 19 that 60% of those capped under the original scheme had been lone parents;

contended in para 21 that, to the extent that the revised cap impacted upon them disproportionately to its impact on others, its impact was justified in the interest of the economic well-being of the UK and of incentivising people to work; and

claimed in para 77 that, in the light of this court’s decision in the first benefit cap case, it had, in relation to the proposed revision of the cap, fully considered what it described as its “obligations” under the UNCRC because the best interests of children overall were promoted when their parents were in work and because work remained the surest route out of poverty.

On 10 September 2015 Gingerbread made representations to the Public Bill Committee of the House of Commons in respect of the proposed revision of the cap. Its policy director said:

On 17 September 2015 Ms Thornberry MP, then the shadow minister of state for employment, tabled before the same committee an amendment to the Bill to the effect that the revised cap should not apply to “persons … responsible for the care of a child aged below two”. The group proposed to be exempted was therefore close to the DA cohorts. But it was not identical in that Ms Thornberry’s amendment appeared to exempt dual-care parents as well as lone parents, although at one point in the discussion she seemed to suggest otherwise. In arguing for her amendment Ms Thornberry suggested that the original cap had overwhelmingly applied to people who were recognised within the benefits system itself as being unable to work. She referred to the evidence which the committee had heard a week earlier, including no doubt that of Gingerbread, and she explained that the narrow exemption which she proposed was for a group that was perhaps the most acutely vulnerable and the least able to change its circumstances. But the committee rejected the amendment by ten votes to five.

Between November 2015 and February 2016 the House of Lords in debates and in committee considered in detail the provision for the revised cap. In summary

Baroness Lister suggested that it was not reasonable to expect a lone parent with a child aged under one to work. She suggested that the government had not properly assessed the best interests of different groups of children pursuant to its obligation under article 3.1 of the UNCRC.

Baroness Manzoor suggested that the lower cap would disproportionately affect single parents with a child aged under five.

Baroness Hollis suggested that lone parents with children aged under three were effectively out of the labour market. She tabled an amendment to exempt carers of children aged under nine months from the cap.

The government’s Equality Analysis dated September 2016 in relation to the 2016 Regulations, like its IA dated August 2016, claimed that the government had taken the UNCRC into account. It stated that it was not in the best interests of children to live in workless households and referred to studies which concluded that children in such households exhibited greater behavioural problems from the age of seven and poorer academic attainment. It recognised that lone parents might find it hard to work as a result of childcare responsibilities but pointed out that measures of mitigation, in particular free childcare and DHPs, had been put in place.

By a narrow margin I am driven to conclude that, in relation to its refusal to amend the 2006 Regulations so as to exempt the appellant cohorts from the revised cap, the government did not breach article 3.1 of the UNCRC in either of the relevant dimensions of its concept of the best interests of a child. The Parliamentary and other materials to which I have referred demonstrate that it did evaluate the likely impact of the revised cap on lone parents with young children; and that it did assess their best interests at a primary level of its overall consideration. This court must impose on itself the discipline not, from its limited perspective, to address whether the government’s evaluation of its impact was questionable; nor whether its assessment of the best interests of young children was unbalanced in favour of perceived long-term advantages for them at the expense of obvious short-term privation.

I am also driven to conclude that the government’s decision to treat the appellant cohorts similarly to all others subjected to the revised cap was not manifestly without reasonable foundation. In this regard, for reasons which I will not rehearse, the DA cohorts have a stronger case than have the DS cohorts; but, again by a narrow margin, even the stronger case fails. The appellants have not entered any substantial challenge to the government’s belief that there are better long-term outcomes for children who live in households in which an adult works. The belief may not represent the surest foundation for the similarity of treatment in relation to the cap; but it is a reasonable foundation, in particular when accompanied by provision for DHPs which are intended on a bespoke basis to address, and which on the evidence are just about adequate in addressing, particular hardship which the similarity of treatment may cause.

There has been no Convention-related discrimination. The appeals must be dismissed.

Had discrimination existed, the court would have proceeded to consider whether to make a declaration that the failure to include the appellant cohorts in the list of exemptions in the 2006 Regulations was incompatible with their Convention rights. A declaration is a discretionary remedy; and to decide whether to exercise the discretion would have precipitated substantial inquiry into the institutional propriety for this court to make a declaration in relation to decisions about entitlement to welfare benefits made by the government in Parliament following protracted debate. But it is this same crucial, if sometimes problematic, concept of institutional propriety which informs the test of justification, generous to the government, of a measure such as that of the revised cap; and it is therefore at that stage that, in relation to such a measure, the concept will usually play its part.

These appeals were rightly brought. The arguments raised in them have been of such weight as to attract this court’s most careful and sympathetic consideration; and they have led two members of the court to enter a powerful dissent from the majority’s dismissal of the appeals. On 12 March 2019, shortly prior to the delivery today of these judgments and long after our hearing of the appeals, the Work and Pensions Committee of the House of Commons published its report on “The Benefit Cap”, 24th Report of Session 2017-19, HC 1477. Although in form a study of the effect of the original as well as of the revised cap, the report inevitably focusses on the current, more severe, effects of the revised cap. It addresses, although in far greater detail, all the factors to which I have referred in paras 22 to 34 above under the heading “Evidence”; as well, of course, as many more relevant factors. In the report the committee calls on the government urgently to conduct a full audit of the policy behind the benefit cap; to reconsider the limits at which benefits are capped; and in particular to disapply the cap to those who, by reference to the conditions attached to the receipt of income support, are not yet expected to look for work. The fact that a committee of the House of Commons is at this present time calling for urgent review of the provisions of the revised cap would in my view have fortified a decision, had the need to make it been reached, that institutional propriety militated against the grant of a declaration of incompatibility at this stage: R (Nicklinson) v Ministry of Justice [2014] UKSC 38, [2015] AC 657, paras 113-118 (Lord Neuberger of Abbotsbury PSC).

I agree with Lord Wilson that the appeal should be dismissed, for the reason that the approach adopted by the government, even if in other respects meeting the tests for discrimination under article 14, was not, as he says, “manifestly without reasonable foundation”. However, having been a member of the majority in the related case of SG, I add some comments on the relationship between the two cases, and some remaining points of difference (or difference of emphasis).

The benefit cap imposes a cap on the total amount of annual welfare benefits that a given household can receive. The legality of the previous scheme under the Welfare Reform Act 2012 was upheld by this court (by a majority) in R (SG) v Secretary of State for Work and Pensions [2015] 1 WLR 1449. Although the scheme was agreed to be discriminatory against women for the purposes of articles 14 and A1P1, it was held to be justified because (in the words of the headnote):

Furthermore the majority held that the case was not strengthened by reference to article 3.1 of the UNCRC. It is convenient again to refer to the summary in the headnote:

The benefit cap legislation was politically controversial and subject to vigorous debate in Parliament, directed to both the principle and the detail. The difficulty posed by the legal issues in that appeal is apparent from the division of opinion within the court, and from the time taken to reach a decision (almost 11 months). However, it must be taken as settling the issues of principle, absent a challenge to the reasoning of the majority, some material change in the relevant legislation or the circumstances of the individual cases, or some new argument of substance which was not addressed.

The present challenge is to the amended scheme under the Welfare Reform and Work Act 2016. By section 8 of the 2016 Act, the applicable cap was reduced from £26,000 per annum for all families to £23,000 for families living in Greater London, and £20,000 for families living outside London (“the Revised Benefit Cap”). A significant change is that the amount of the cap is embodied in primary legislation, rather than regulations, as is the list of benefits to which it applies, including child benefits and child tax credit (section 96(10)). It is rightly not suggested that these aspects of scheme, as now incorporated in primary legislation, are in themselves open to review or incompatible with the Convention. To that extent the scope for challenge has been reduced.

The 2012 Act gave the Secretary of State the power to make detailed provision for the implementation of a benefit cap by way of regulations. Under section 96(4)(c) of the 2012 Act, this includes a discretion to make exceptions to the application of the cap. The Benefit Cap (Housing Benefit) Regulations 2012 (SI 2012/2994) provide, inter alia, that adults who are entitled to “working tax credit” are not subject to the benefit cap. In the case of a lone parent, receipt of working tax credit requires that he or she engages in work for at least 16 hours per week. It is the regulations which are the focus of the present attack. In summary the appellants argue that their parental responsibilities, combined with the lack of adequate care support, make it in effect impracticable for them to achieve the 16 hours of work necessary to qualify for exemption. It is argued that failure to make an appropriate exception for them under the regulations involved unjustified discrimination contrary to article 14 of the ECHR.

Apart from the change in legislation, there appear to be three main differences from the arguments as presented in the earlier cases:

Article 8 Particular emphasis has been placed on article 8 of the Convention (rather than article 1 of Protocol 1 (“A1P1”) which was the main focus of attention in SG), and its relevance to the “best interests” test under the UNCRC;

Status The focus has shifted to the so-called Thlimmenos principle, and the groups allegedly discriminated against have been recalibrated and re-defined in various ways: (DS appellants) (i) lone parents, (ii) lone parents with children under the age of five (iii) children of parents in groups (i) or (ii); (DA appellants) (iv) lone parents with children under two (v) children of such parents.

Test for justification It is argued that in the light of more recent Supreme Court authority, the “manifestly without reasonable foundation” criterion applied in SG, has been superseded by a broader “fair balance” test.

I will consider each of these points in turn before explaining my conclusions in the present appeals.

As I noted in SG (para 99) article 8 had been mentioned by Mr Wise in his printed case, not as a free-standing claim, but as an alternative route into article 14, or as supporting his “best interests” claim in respect of the children under article 3.1 of the UNCRC. I noted that article 8 was not relied on by Mr Drabble QC, then appearing for the Child Poverty Action Group. I was not at that time persuaded that “either of Mr Wise’s formulations adds anything of substance to the claim based on A1P1”. It may be in retrospect that we should have given more attention to this aspect of Mr Wise’s submissions. In any event, there is no doubt that the main weight of the argument at that time, and the reasoning of the majority, were directed to A1P1 rather than article 8.

Lord Hughes in particular drew a clear distinction between the two in the particular context of the “best interests” principle under article 3.1 of the UNCRC:

Accordingly the present appeal has required us to look in more detail at the application of article 8. As to the application of article 14 in connection with article 8, it is unnecessary to repeat Lord Wilson’s review of the relevant Strasbourg authorities under his issue 1. In agreement with him I am satisfied that the present claims fall within “the ambit” of article 8 so as to engage the issue of discrimination under article 14. I also agree with him that in that context the “best interests” principle under article 3.1 is potentially relevant. I have more difficulty with the issue of “status” to which I now turn.

Although the Thlimmenos principle is now well-established, it does not in my view materially change the nature of the inquiry from that undertaken in SG. In particular it does not diminish the need under article 14 to show that the alleged discrimination arose from a relevant “status”, and to identify a relevant “comparator” with whose treatment that of the claimant group can be compared (Lord Wilson’s issues 2 and 4).

In Thlimmenos v Greece the applicant was a Jehovah’s Witness who had been convicted of insubordination under the Military Criminal Code for refusing to wear a military uniform at a time of general mobilisation. He was subsequently refused appointment as a Chartered Accountant under rules which excluded those convicted of serious crimes. He argued that the lack of an appropriate exception for those whose conviction was due to religious considerations constituted unlawful discrimination under article 14 taken with article 9 of the Convention.

The argument was accepted by the Grand Chamber. Having noted that article 14 had hitherto been applied to differential treatment of persons in analogous situations without objective and reasonable justification, the court continued:

Although the court’s formulation of the principle does not refer to “status” as a criterion, it is apparent from the preceding discussion that this point was not in issue. In that respect the applicant’s argument as recorded by the court relied on his position as a Jehovah’s Witness:

The Thlimmenos formulation has been often repeated and is not in doubt. However, there are few illustrations of its practical scope and application. An example is EB v Austria (Application No 31913/07, judgment of November 7, 2013), in which the First Section found that the principle required an exception to the general rule that convictions remained on the person’s record even if the offence in question had since been abolished. In that case, there was a violation of article 14 taken with article 8 as convictions for homosexual acts, later found to be incompatible with the ECHR, remained on the applicants’ criminal records. We were referred to no Strasbourg case in which the principle has been applied in the context of social welfare legislation such as is in issue in this case. Although there is no reason to exclude its operation in this context, the absence of successful cases in Strasbourg may reflect the court’s recognition in this context of the “need for national rules to be framed in broad terms” (SG para 15 per Lord Reed citing Carson v United Kingdom (2010) 51 EHRR 13 para 62), and the consequent difficulty of challenging the treatment of particular groups.

I must accept (as Lord Wilson says: para 39) that in R (Stott) v Secretary of State for Justice the majority of this court adopted a relatively broad view of the concept of “status”. On that basis I would agree that “lone parents” can properly be regarded as having a status within the Thlimmenos principle. In agreement with Lord Hodge, I am much more doubtful as to the appropriateness of the other narrower forms of status relied on in this case. In particular I find it hard to see any basis for defining the parents and the children as distinct groups; the adverse effects are on the families, in which the interests of parents and children are jointly affected. However, in the absence of any directly relevant Strasbourg authority on these points, it is difficult to reach a concluded view. Like Lord Hodge I am content to assume for present purposes that the “status” requirement is satisfied in respect of each such group.

The relevant issues therefore are whether those groups or sub-groups are sufficiently different from other comparable groups to have required separate treatment under the Thlimmenos principle to avoid interference with their article 8 rights, and whether a failure in that regard can be justified.

The argument that a less demanding test should be applied than “manifestly without reasonable foundation” (or its hard-to-escape acronym “MWRF”) was most fully articulated by Mr Wise QC for DA. For the reasons given by Lord Wilson (issue 7) I agree with him that this argument must be rejected, and that the application of the MWRF should be regarded as beyond “future doubt”. However, since this view is not accepted by all the members of the court, I feel it necessary to add some comments of my own on Mr Wise’s arguments.

He started from the four-stage approach as summarised by Lady Hale in R (Tigere) v Secretary of State for Business, Innovation and Skills [2015] UKSC 57; [2015] 1 WLR 3820, para 33:

He accepted that the MWRF test was adopted by this court in Humphreys v Her Majesty’s Revenue and Customs Comrs [2012] UKSC 18; [2012] 1 WLR 1545 (alleged discrimination in respect of state benefits), following the ECHR decision in Stec v United Kingdom (2006) 43 EHRR 1017, and in other more recent cases. However, in those cases, as he submitted (in his written case) -

With respect to those members of the court who think otherwise, it is clear in my view that the MWRF test remains the appropriate test in the present context. There is nothing in the later cases to support a departure from the position, as accepted by all parties, and adopted by the court in the SG case. It is useful to begin by reference to what was said by Lady Hale (with the agreement of the rest of the court) in the Humphreys case itself:

As that passage shows, authority at the highest level in this country for the application of the MWRF test goes back at least to the House of Lords in RJM. Also noteworthy is her reference to the distinction drawn by the House of Lords between the “core” grounds such as race and sex, and other grounds; and her acceptance that, even in the core context of sex discrimination, the “normally strict test” for justification “gives way” to the MWRF test in the context of state benefits. In the SG case itself the discrimination was said to be against women, and thus within one of the core grounds. As one moves further away from those concepts to the more distant groups identified in the present case, there is still less reason to depart from the MWRF approach.

SG was argued in April 2014 but not decided until March 2015. As already noted, there was no disagreement between the parties as to the application of the MWRF test. By that time the decision in the Welsh Asbestos case had been given (and was mentioned by Lady Hale: para 209); but it was not treated by the majority, or indeed the parties, as requiring any qualification of the MWRF test. Nor was there any such departure or qualification in the Tigere case itself (decided in July 2015). The case concerned discrimination in relation to eligibility for student loans. There was a disagreement between the members of the court over the appropriate test on the facts of that case, but not on the correctness of the MWRF as applied in the cases of Humphreys or SG. Lords Sumption and Reed would have applied the MWRF test. Lord Hughes concluded that the appeal should be allowed whatever the test. Lady Hale (paras 27-29, with the agreement of Lord Kerr) referred to those judgments without adverse comment, saying “… education is rather different”. She went on to cite the Strasbourg decision in Ponomaryov v Bulgaria (2011) 59 EHRR 799, including in particular to observation of the court that “unlike some other public services, education is a right that enjoys direct protection under the Convention ...”.

As Lord Wilson says (para 63), the issue was in any event put beyond reasonable argument by the seven-justice court (including Lady Hale and Lord Mance) in the “bedroom tax” case (R (MA) v Secretary of State for Work and Pensions). Giving the leading judgment (with the agreement on this point of all members of the court), Lord Toulson noted the submission that, because in Humphreys the unsuccessful appellant had not argued for anything other than the Stec test, it was appropriate to ask whether there was good reason to depart from what Lady Hale had said (para 31). As he then pointed out (para 32):

In her judgment in the present case, as I understand it, Lady Hale does not seek to question the application of the MWRF principle in these appeals, but suggests that the court may need to return to it in the future. I cannot with respect agree. She accepts that the decision in MA was reached following “a wholesale attack” on the MWRF principle, but observes that there was no discussion of a “more nuanced” approach along the lines suggested by Lord Mance in the Welsh Asbestos case. I do not see that this in any way diminishes the authority of the decision. It is to be noted that, in spite of the presence of Lord Mance, and although the Welsh Asbestos case was included in the list of authorities cited, neither he nor anyone else seems to have regarded it as relevant in that context. That may well have been because the context in which the issue was considered was quite different from MA (and from the present case): not social security benefits, but compensation for asbestos-related disease; and not article 14 discrimination, but interference with property rights under A1P1. Indeed in Welsh Asbestos there had been no reference to Stec or Humphreys in either the judgment or in the submissions.

Lord Kerr goes further and would hold, in agreement with Mr Wise’s submission, that the MWRF test should not be applied to the final stage of the proportionality analysis. Although he does not in terms explain how he feels able to disregard the authority of MA, he emphasises that the technique applied to that question by the national court is to be distinguished from that applied in Strasbourg at the supra-national level. However, the fact that the Strasbourg court uses the MWRF test when applying the margin of appreciation and that the same margin of appreciation does not necessarily apply at the national level does not entail that domestic courts cannot also use the MWRF test. It is being used as a means of allowing the political branches of the constitution an appropriately generous measure of leeway when assessing the proportionality of measures concerning economic and social policy. The seven-justice decision in MA surely settled the point for the foreseeable future.

In conclusion I adopt with respect what was said by Sir Patrick Elias in the Court of Appeal in DA (which related solely to the group said to be constituted by lone parents with children aged under two):

Although the number of possible groups has been extended in the appeals as they have come to this court, the thrust of that passage remains valid. It is necessary to distinguish between the general impact of the cap, which is undoubtedly harsh, but is inherent in the scheme as approved by Parliament, and particular effects on an identifiable group which can properly be the subject of a distinct claim under article 14.

Applying that approach, I ask whether there are factors in the present cases which require the court to reach a different overall conclusion from that reached in SG. I have noted that in some respects the task facing the appellants is more difficult. The amount of the cap, and the benefits to which it applies (including child benefit) are enshrined in primary legislation, which is admittedly not open to challenge. Although I have accepted that the various groups identified by the claimants can be regarded as meeting the “status” requirement for the purposes of article 14, they are far from the “core” grounds to which special protection is given under that article, and in relation to which the court should be especially slow to substitute its view for that of the executive (see para 113 above, citing RJM para 57).

On the other side, I have accepted that, in contrast to the position in SG, the claimants are able to pray in aid the best interests principle under article 3.1 of the UNCRC. However, in that respect the extracts from the Parliamentary debates quoted by Lord Wilson show that careful consideration was given, not only by the executive, but also by Parliament, to the extent to which further exceptions should be enacted, and in particular to the interests of the children potentially affected. I agree with him that it has not been shown that the failure to enact further exemptions involved any breach of that principle. My contrary conclusion on that issue in SG was narrowly based on the deficiencies in the Secretary of State’s evidence on this aspect (paras 110-112, 127-128), and has no relevance to the present appeals. Overall I agree with Lord Wilson that the approach ultimately adopted by the executive, with the support of Parliament, was not manifestly without reasonable foundation, and that the appeals must accordingly be dismissed.

As a final comment, and without disrespect for the care and skill with which the cases have been presented to the court, I observe that the dangers of departing from the restrictive approach laid down by Lord Toulson in MA are amply demonstrated by the experience of this appeal. We have been faced with detailed submissions based on conflicting factual and statistical evidence, much of it produced for the first time in this court. Some of this evidence has come in support of submissions from interveners. Their experience of the practical implications of the legal issues can be of great value, but the court must be careful to ensure that such interventions do not lead to the introduction of new evidence which has not been fully tested, and which cannot be properly tested within the limitations of this court’s proper function. At times it has seemed as though the court were being invited to take on the task of a Parliamentary Select Committee, undertaking a review of the policy and factual basis of the legislation. That is not our role.

I agree with Lord Wilson that the appeals should be dismissed for the reasons which he gives. I wish to add only one qualification to my agreement and that relates to the question of status. In this regard I share the doubts which Lord Carnwath expresses on this issue in para 108 of his judgment.

I also agree with Lord Carnwath’s view on justification (the “MWRF test”) in paras 110-118 of his judgment, which tallies with that of Lord Wilson. As Lord Kerr states, the precise reason why the ECtHR adopted the MWRF test does not apply to the domestic court. But it is open to a domestic court to adopt that test in relation to socio-economic policy decisions of the executive in recognition of the institutional constraints on and the constitutional role of the judiciary. That is what both the House of Lords and the Supreme Court have done in the cases of R (RJM), Humphreys, SG and MA, as Lord Carnwath demonstrates in his judgment.

For the purposes of these appeals I am content to assume that each of the claimants has the required status to mount a challenge under article 14 of ECHR. But this appeal, like the appeal which this court heard in R (Stott) v Secretary of State for Justice [2018] UKSC 59; [2018] 3 WLR 1831, raises questions on the boundaries of “other status” in article 14, a subject on which there is, as yet, little clarity. Some may argue that the requirement of status is not an important hurdle for a claimant to overcome and that the Convention requires the state to justify any failure to treat differently people whose situation is relevantly different. But as national rules on social security benefits are required to be expressed in broad terms which will affect different people differently, the lack of clarity as to the entitlement of groups and sub-groups to challenge is a mischief. I do not therefore wish to endorse the view that each of the cohorts of claimants has the necessary status.

In these appeals, the DA appellants assert a status as lone parents of children aged under two and as children aged under two of lone parents. The adult DS appellants assert a status as either lone parents or, as a fall back, as lone parents with children aged under five. The children who are DS appellants assert a status as children of lone parents or, as a fall back, as children aged under five of lone parents. Thus, the court faces the question whether lone parents of children of any age, lone parents of children aged under five, lone parents with children aged under two, and their respective children each enjoy a separate status under article 14.

When one considers an arrangement which requires someone in a household to work for a minimum number of hours per week in order to escape the benefit cap, it is not difficult to see that a lone parent household is in a different situation from a two parent household. Indeed, the government recognizes that difference by setting a lower minimum number of hours of work per week for the lone parent household. But, as Lord Wilson points out, there will be many within this cohort of lone parents whose children are all of school age. It is, as he states (para 45), a simpler task for the government to justify the requirement that a lone parent with children all of school age should work at least 16 hours per week to escape the benefit cap than it is to justify that requirement for lone parents with children under school age.

When children are under school age, a lone parent’s ability to work at least 16 hours per week is dependent on the availability of the support of others in child care. Lone parents with babies and toddlers can be expected to have greater difficulty in working those hours without such childcare support. But where does one draw the line or lines? Is there not a sliding scale? The government’s rules on the availability of financial support for childcare supports the view of the policy director of Gingerbread: “the younger the child is when the parent is capped, the harder it is for them to get into work” (see Lord Wilson’s judgment para 83 above). But the age of the youngest child is not the only variable which influences how hard it is for a lone parent to obtain work. The availability of part-time work, the proximity of family or friends who can provide child care, and the availability and funding of childcare facilities are equally important variables. Ill-health and other adverse circumstances of the parent or children in a large family may militate against a lone parent’s ability to work, regardless of the age of the youngest child.

Looking more broadly, the difficulty in escaping from the benefit cap is only one feature of the cap. Other circumstances are also relevant to the burden which the cap imposes on parents. As Sir Patrick Elias stated in his leading judgment in the Court of Appeal (para 105), there is no linear relationship between the age of the children and the financial impact on families caused by the cap. The greater the number of dependent children of whatever age there is within a household, the more mouths there are to feed and larger is the accommodation that the family needs.

I am left with some doubt as to whether the age of a lone parent’s youngest child is the basis for giving the parent and child a status for the purpose of article 14 in the circumstances of this challenge. The people with the strongest case for having their circumstances recognised as giving rise to a status, it seems to me, are the DA cohort of lone parents with children aged under two and those children, having regard both to the degree of dependence of the child and the manner in which the government recognises that dependence both in the non-provision of free child care to most two year olds (para 28 above) and in the conditions set for the receipt of income support which are graduated by reference to the age of the child (para 30 above). But I am content to leave the question of status to future dialogue with the ECtHR.

It is indeed regrettable that there is a variety of opinions among the judges who have considered these cases and accordingly that it has taken this court so long to produce its judgment: regrettable but not at all surprising. These are cases about equality and equality is the most complicated and difficult of all the fundamental rights, even without the delicate context of entitlement to welfare benefits. A professional lifetime of struggling with equality issues has persuaded me that some degree of complexity is inevitable and we should not apologise for it. The law may be complicated and sometimes difficult to apply but for the most part it does not lack clarity. There is no difference of opinion between Lord Wilson and me as to the legal principles applicable: we disagree only on the application of the principle of justification to the facts of these cases.

The delicacy arises because these are cases about equality in an area, not principally of social policy, but of economic policy. Constitutionally, economic policies are decided by those organs of government which are directly accountable to the people. The courts cannot make those decisions for them. But that does not mean that the courts have no role to play. In a constitution which respects and protects fundamental rights, it is the role of the courts to protect individuals from unjustified discrimination in the enjoyment of those fundamental rights. There are no “no go” areas. The courts might very well have declared that denying certain widows’ benefits to widowers was incompatible with the Convention rights, had the Strasbourg court not got there first: see Willis v United Kingdom (2002) 35 EHRR 547. More recently, this court has declared the denial of widowed parent’s allowance to a surviving parent who was not married to the deceased parent of their children incompatible with the Convention rights: see In re McLaughlin [2018] 1 WLR 4250. But those examples are more clear-cut than these. Nor do they mean that the courts will not recognise that the government is both constitutionally and institutionally more competent than the courts to make the delicate judgments involved: see, for example, R (Hooper) v Secretary of State for Work and Pensions [2005] 1 WLR 1681.

The argument before us now is very different from the argument which was before us in R (SG) v Secretary of State for Work and Pensions [2016] 1 WLR 1449, although that too concerned the benefit cap, albeit in its original and less draconian version. In SG the complaint was of indirect discrimination against lone parent women. It was indirect because the benefit cap applied equally to all lone parents, men and women. But the government acknowledged that it had a disproportionate impact upon women because the overwhelming majority of lone parents are women. The debate was about whether it could be justified and about the relevance of the United Kingdom’s international obligations under the United Nations Convention on the Rights of the Child (UNCRC) to that question. A majority of this court concluded that the government had not complied with its obligation, under article 3.1 of UNCRC, to treat the best interests of the children concerned as a primary consideration. But a majority also concluded that this was not relevant to whether the indirect discrimination against women was justified. Although I disagreed with that conclusion, I found it completely understandable. The children of lone parents were hit equally hard by the benefit cap whether their parents were men or women. The relevance of their interests to the alleged sex discrimination was therefore questionable. Indeed, as I had said in Humphreys v Revenue and Customs Comrs [2012] 1 WLR 1545, a case alleging indirect sex discrimination in the rules governing entitlement to child benefit, which did not allow the benefit to be split between shared carers (para 20):

In these cases, the basic complaints are of discrimination between, on the one hand, lone parents and their children and, on the other hand, dual parents and their children. In DA, it is narrowed down to a complaint of discrimination against lone parents with a child or children under the age of two and their children, whether compared with dual parents with children under two or other lone parents. Both complaints are easier to grasp than the complaints in SG.

Article 14 of the European Convention on Human Rights (ECHR), as is well known, provides:

There is nowadays no doubt that entitlement to state benefits, even non-contributory means-tested benefits, is property for the purpose of article 1 of the First Protocol (A1P1), which protects property rights. Indeed, the benefit cap can be seen as a species of deprivation of property, as it takes away rights which the claimants would otherwise have. But, as Lord Wilson explains (para 36), benefits which enable a family to enjoy “a home life underpinned by a degree of stability, practical as well as emotional, and thus the financial resources adequate to meet basic needs, in particular for accommodation, warmth, food and clothing” are clearly one of the ways (“modalities”) whereby the state manifests its respect for family life and therefore fall within the ambit of article 8 (see Petrovic v Austria (2001) 33 EHRR 14 and Okpisz v Germany (2006) 42 EHRR 32). That we are concerned here, not only with the right to property, but also with the right to respect for family life is clearly relevant to the issue of justification.

The government’s argument that, because the claimants are women, who already have a status under article 14, they should not seek to shoehorn themselves into some other status (see para 39 above) is clearly unsustainable. Men also have a status under article 14, but they often qualify for some other status, such as being married or unmarried (In re G (Adoption: Unmarried Couple [2009] 1 AC 173), disabled or not disabled (R (RJM) v Secretary of State for Work and Pensions [2009] 1 AC 311) or serving an extended sentence of imprisonment or some other sentence (R (Stott) v Secretary of State for Justice [2018] 3 WLR 1831). In any event, the DA claimants are not complaining of discrimination because they are women. A male lone parent could have brought exactly the same complaint. And that also applies to the DS claimants (although they also revive the indirect sex discrimination claim in SG). Lone parent is clearly a status within the meaning of article 14. And I agree with Lord Wilson and Lord Kerr that it can be sub-divided according to the ages of the children, and in particular that having a child or children under compulsory school age is obviously a status for this purpose, just as being a disabled child who needed more than 84 days’ hospital in-patient care was a status in Mathieson v Secretary of State for Work and Pensions [2015] 1 WLR 3250 and indeed being a particular type of prisoner was a status in R (Stott) v Secretary of State for Justice, above.

I agree with Lord Wilson and Lord Kerr that the natural way to formulate the DA complaint is that these lone parents, and their children, should have been treated differently from other lone parents, and from dual parent families, because their situation is relevantly different from the situation of other families who are subject to the benefit cap: that is, discrimination within the meaning of Thlimmenos v Greece (2000) 31 EHRR 12.

I also agree that the natural way to formulate the DS complaint is of Thlimmenos-type discrimination, whether it extends to lone parents of children of up to school-leaving age or only to lone parents of children under compulsory school age. As already explained, I think that any attempt to formulate the DS complaint as one of indirect discrimination against women would run into the same difficulties as were encountered in SG.

In both cases, the natural comparator is a dual parent family with children of the same age. But this is complicated because dual parent families are, in at least two respects, less favourably treated than lone parent families: they are subject to the same cap on their benefits although they have one more adult mouth to feed; and they can only escape if between them they work outside the home for a total of 24 hours a week. Although this does not necessarily present insuperable problems, these do not arise if the DS complaints are limited to those with a child or children under five. Then both the DA and the DS claimants can compare themselves with lone parents who only have older children.

It can immediately be seen that the situation of these claimants is relevantly very different from the situation of lone parents with children of school age. Lord Wilson has summarised this with great clarity in para 51 above. I would lay particular stress in what he says at para 51(a), which is worth repeating:

It is dangerous for a judge to indulge in moral indignation but few mothers (and indeed few lone fathers) who have chosen to work also outside the home while their children are very young can have escaped being made to feel guilty that they may have been harming their children’s healthy development by doing so. We were brought up on John Bowlby’s classic work, Child Care and the Growth of Love, the foundation of modern attachment theory. Children need to form stable and healthy attachments early in life in order to be able to lead healthy lives and form stable attachments of their own in the future. The foundation for this is stable, consistent and loving care from a parent or parents (or parent-substitutes) who have bonded early with the child. No-one who has sat as a judge in the family courts can have escaped hearing constant evidence about the risks of significant harm to children who are denied such healthy attachments.

This is not to say that children cannot also thrive if their parents do go out to work. Such work may bring psychological as well as financial benefits to their parents, as well as to society, and this may also benefit their children. But for this the children need good, stable and consistent alternative care arrangements, preferably in familiar surroundings: children develop attachments to places as well as people. Such arrangements are in short supply and very expensive. The availability of help towards the cost of up to 15 hours’ child care for some of these children does not necessarily fit this bill. The government itself has recognised that parents of very young children should not be obliged to seek work outside the home, both by the conditions they have set for eligibility for state benefits and by the limits they have set for free child care. The psychological risks to children whose lone parents are obliged to work outside the home in order that their children may have enough to live on, whether or not this is in their children’s best interests, have to be set against the psychological risks to children who grow up in benefit-claiming families, risks to which the government has attached so much importance.

Of course, those risks will be less if the parent can find suitable work as well as suitable child care. She may be lucky enough, for example, to find some evening shifts in a very local supermarket and have a willing grandparent or neighbour to look after the children while she does so. But any lone parent who has small children will face considerable difficulties in finding suitable work which will fit in with her child care arrangements and also, in many cases, with her commitments to her other children.

The one matter on which the law may be open to debate relates to the standard by which the government’s justification for discriminatory measures such as this is to be judged. In Humphreys v Revenue and Customs Comrs [2012] 1 WLR 1545, this court, in a judgment of mine with which the other Justices agreed, adopted the “manifestly without reasonable foundation” standard derived from the Strasbourg Grand Chamber decision in Stec v United Kingdom (2006) 43 EHRR 47, albeit with the qualification that this did not mean that the justification advanced should escape “careful scrutiny” (para 22). But that test was not disputed in Humphreys, any more than it had been in R (RJM) v Secretary of State for Work and Pensions [2009] 1 AC 311.

However, Lord Kerr is surely right to question whether the test which the Strasbourg court will apply in matters of socio-economic policy should also be applied by a domestic court. The Strasbourg court applies that test, not because it is necessarily the proper test of proportionality in this area, but because it will accord a “wide margin of appreciation” to the “national authorities” in deciding what is in the public interest on social or economic grounds. The national authorities are better able to judge this because of their “direct knowledge of their society and its needs” (see Stec, para 52). It does not follow that national courts should accord a similarly wide discretion to national governments (or even Parliaments). The margin of appreciation is a concept applied by the Strasbourg court as part of the doctrine of subsidiarity. The standard by which national courts should judge the measures taken by national governments is a matter for their own constitutional arrangements.

Not only that, it has been noted that, in Stec, the Grand Chamber cited James v United Kingdom (1986) 8 EHRR 123 as authority for its “manifestly without reasonable foundation” standard. But in James, it is fairly clear that the Strasbourg court drew a distinction between two questions: first, was the measure “in the public interest” for the purpose of A1P1 (or, in article 8 terms, does it pursue a legitimate aim); and second, was there a reasonable relationship of proportionality between the means employed and the aim sought to be realised. This latter requirement had been expressed in Sporrong and Lönnroth v Sweden (1982) 5 EHRR 35, at para 69, as “whether a fair balance was struck between the demands of the general interest of the community and the requirements of the protection of the individual’s fundamental rights” (see James, para 50). The “manifestly without reasonable foundation” standard was applied to the first but not the second question.

When discussing (albeit strictly obiter) whether the imposition of charges for NHS treatment for asbestos related diseases caused by employers’ breach of duty was compatible with the A1P1 rights of employers and their insurers, in In re Recovery of Medical Costs for Asbestos Diseases (Wales) Bill [2015] AC 1016, para 51, Lord Mance noted that our domestic law had also drawn a distinction between whether the aims are legitimate and whether a fair balance had been struck. Both Lord Hope and Lord Reed had adopted this approach in AXA General Insurance Ltd v HM Advocate [2012] 1 AC 868. Even in relation to the “fair balance” question, of course, “domestic courts cannot act as primary decision makers and principles of institutional competence and respect indicate that they must attach appropriate weight to the informed legislative choices at each stage of the Convention analysis: see the AXA case, … para 131” (Asbestos, para 54).

I pause only to note that if Lord Hope, Lord Mance and Lord Reed were correct in separating the issues of legitimate aim and fair balance in A1P1 cases, and applying a different standard to each, it would be wrong to apply a different approach to those same questions when they come up in the context of discrimination in the enjoyment of the right to respect for family life. The principles applicable when, say, insurance companies challenge interferences with their property rights should not be more favourable to them than the principles applicable when children challenge discrimination in their right to respect for their family lives.

In R (MA) v Secretary of State for Work and Pensions [2016] 1 WLR 4550, the so-called “bedroom tax” case, a wholesale attack was mounted upon the “manifestly without reasonable foundation” test. The attack was rejected in favour of the continued application of the Humphreys approach (para 38). There was no discussion of a more nuanced approach along the lines suggested by Lord Mance in Asbestos. Even applying Humphreys the removal of the spare room subsidy was found to be without justification in two respects. I agreed with that judgment (para 81).

It seems to me that the court may well have to return to this difficult question in another context at some point in the future. But this is neither the case nor the context to do so. Nor is it necessary. The government has put forward three aims for the revised benefit cap: (i) fairness as between those in work and those on benefits, in that those in work should always be better off than those who are not, and with it the maintenance of public confidence in the benefits system; (ii) fiscal savings; and (iii) incentivising work outside the home. These are indeed legitimate aims, whatever the standard by which they are judged.

However, it is also an essential element in justification that the measures adopted should be rationally related to their legitimate aims (see, among many, Lord Reed in Bank Mellat v HM Treasury (No 2) [2014] AC 700, para 76). This is another way of putting the requirement in European Union law that a measure be “suitable” to achieving its aims. It seems to me that it has been comprehensively demonstrated by the mass of evidence before the trial judge in DA, Collins J, and summarised by Lord Wilson and Lord Kerr, that the revised benefit cap is not suitable to achieving any of its declared aims. It does not achieve fairness as between those in work and those on benefits, because those in work will always be better off than those who are not. The fiscal savings are very small and liable to be offset by increased costs in other budgets. These include discretionary housing payments and the cost of housing and rehousing families made homeless as a result of the cap, as well as the costs resulting from the harm done to children by the disruption to their lives and education, as well as by living in poverty, in their early years: the fact that these costs will mainly fall upon local authorities rather than central government makes no difference in principle. There will be other costs if the lone parent is driven to take work outside the home, but it has not been shown that the benefit cap has this effect on this particular group of lone parents.

But even if it could be shown that the benefit cap does have some effect in fiscal savings overall and inducing lone parents of young children to work outside the home, the question of a fair balance between the benefits to the community and the detriment to the children and their parents would still arise. The government is under an obligation in international law to treat the best interests of the children concerned as “a first priority”. It has been held on several occasions that whether it has fulfilled that obligation is relevant to whether it has acted compatibly with the Convention rights of the children concerned: see ZH (Tanzania) v Secretary of State for the Home Department [2011] 2 AC 166, H (H) v Deputy Prosecutor of the Italian Republic, Genoa [2013] 1 AC 338, Zoumbas v Secretary of State for the Home Department [2013] 1 WLR 3690.

In showing that a fair balance has been struck, it is not enough for the government to show that it was aware of the concerns raised by many in and outside Parliament about the effect of the revised benefit cap on the welfare of children in lone parent families. Awareness is not the same as taking the best interests of those children seriously into account. Even taking them into account is not the same as giving them first priority which is an intrinsic part of striking a fair balance where children’s rights are concerned.

Striking a fair balance would have set the very limited benefits to the public interest against the damage done to the family lives of young children and their lone parents if either their parents are forced to work outside the home in order to have enough for themselves and their children to live on or they are unable or unwilling to work outside the home and are thus forced to attempt to live on less than the state has decided that they need. In particular, there is little or no evidence that proper account has been taken of the risks of psychological harm to very young children if they are separated from their primary carers, or the multiple risks to the health, development and life chances of children living in poverty in their early years. There is little or no evidence that these very real and well-documented risks have been fairly balanced against the much more speculative risks of spending those very early years in a household dependent on welfare benefits - we are talking here of children who are below compulsory school age, whose understanding of where the money to live on comes from will be limited, although of course there may be older children in the same household. Once all the children are of school age, there will be ample incentive for their parents to try and find work outside the home if they can.

Therefore, with the greatest respect for the institutional competence of the government as primary decision-maker in these matters, this seems to me a clear case where the weight of the evidence shows that a fair balance has not been struck between the interests of the community and the interests of the children concerned and their parents. I would therefore allow the appeals and make the declaration made by Collins J in DA, amended to include families with children under compulsory school age in DS, as follows:

There is much in the judgment of Lord Wilson with which I completely agree. Indeed, the areas of disagreement between us, although important, are relatively few. It is right that I should express my admiration for his judgment and my indebtedness to Lord Wilson for his distillation of the many complex issues that arise in this difficult appeal.

In particular, I agree with Lord Wilson, for the reasons which he gives in paras 35 and 36 of his judgment, that the appellants’ claims fall within the ambit of article 8 of ECHR; that all the appellants have the requisite status to advance claims under article 14 - paras 38 and 39; and that the natural way to analyse the complaints of the DA appellants is as discrimination of the type explained in the Thlimmenos case - para 40. I also agree that, in relation to the DS appellants, the court should address the fall-back complaint (ie that relating to lone parents of children under five) and should again do so as a Thlimmenos complaint - paras 44 and 45.

In paras 46 and 47 of his judgment, Lord Wilson explains why he considers that the natural comparators to the cohorts whom the appellants represent are all others subject to the revised benefit cap. I agree with his analysis, and I also agree with his conclusion that it is open to the appellants, in advancing the argument that there has been an objectionable similarity of treatment of the DA and DS cohorts and all who have been subject to the cap, to draw particular attention to the marked difference between their situation and those in sub-groups (a) and (b), namely dual-care parents with a child under two or under five, and lone parents without a child under two or under five. Treating the DA and DS cohorts in precisely the same way as the members of those groups when there are significant differences in their respective circumstances sounds directly on the issue of justification.

It follows from what I have said in the previous paragraph that I agree with Lord Wilson that there is clear prima facie evidence that the appellants are in a relevantly different situation from others who are subject to the revised benefit cap - para 51 of his judgment. The factors identified in sub-paras (e) and (f) of para 51 are of especial significance.

As Lord Wilson has pointed out in para 53 of his judgment, the authoritative statement on what requires to be justified is found in the speech of Lord Bingham in A v Secretary of State for the Home Department [2004] UKHL 56; [2005] 2 AC 68, para 68. What requires to be justified is the difference in treatment between one group and another. That requirement translates in the present case to justification of the same treatment to two disparate groups where their circumstances differ to the extent that they plainly call for differential treatment. Specifically, what must be justified here is the decision not to make provision for exemption of the DA and DS cohorts by amending the 2006 Regulations.

In para 55, Lord Wilson adverts to “two different paths” which this court has followed in its pursuit of the proper test against which justification is to be measured where what is involved is an economic measure “introduced by the democratically empowered arms of the state.” Lord Wilson considers that this “duality” has been unhelpful and expresses regret for having contributed to it. In my view, there is no reason for regret and, while the divergence of opinion on this issue might be considered unfortunate, it is, I am afraid, unavoidable and cannot be swept away.

The divergence centres on the question of the use of the formula commonly known as “manifestly without reasonable foundation”, when examining the proportionality of measures devised by government or the legislature in the fields of economic or social policy. This principle, if it is appropriate to describe it as such, is the creature of the European Court of Human Rights (ECtHR). Its provenance is the margin of appreciation which ECtHR accords to decisions of national authorities in the fields of economic and social policy particularly. Thus, as Lord Wilson observed in para 58, in the cases of James and Carson the Strasbourg court held that respect should be shown to the national legislature’s decision on matters of public interest when devising economic or social measures unless it was manifestly without reasonable foundation. It is significant that, as Lord Wilson explained, what he described as this more benign approach “flowed from the margin of appreciation”. The manifestly without reasonable foundation formula should be recognised as a fundamental element of the margin of appreciation doctrine, therefore.

This much is clear from the decision of the Grand Chamber in Stec v United Kingdom (2006) 43 EHRR 47. In that case ECtHR endorsed the manifestly without reasonable foundation approach in assessing whether a measure of economic policy, said to offend article 14 of the Convention, was justified. But this was expressly linked to the application of the margin of appreciation principle. At para 52 the court explained the reason for its reluctance to interfere in this way:

The interconnectedness between the manifestly without reasonable foundation rubric and the margin of appreciation doctrine is therefore clear. On that account, a failure to acknowledge that rubric as an element or sub-set of the margin of appreciation principle can lead to an unwitting importation of a quasi-margin of appreciation approach into the national courts’ consideration of the proportionality of a measure. This is impermissible even in the fields of economic or social policy. There may have been a tendency to do precisely that, however, in some earlier decisions of this court. Thus, for instance, in MA [2016] UKSC 58; [2016] 1 WLR 4550, para 32, Lord Toulson adopted for the purposes of national courts’ review the standard prescribed by Strasbourg when he said:

But there is plenty of authority which acknowledges that measures falling within the United Kingdom’s margin of appreciation, when viewed from the supra-national perspective of ECtHR, will not necessarily survive judicial scrutiny on the national stage. In In re Recovery of Medical Costs for Asbestos Diseases (Wales) Bill [2015] UKSC 3; [2015] AC 1016, para 54 Lord Mance said:

Lord Neuberger picked up this theme in Nicklinson [2014] UKSC 38; [2015] AC 657 where, at para 74, he said:

The importation of the test “manifestly without reasonable foundation” to all aspects of the national court’s consideration of proportionality imperils the proper discharge of its duty. This was a technique devised by the Strasbourg court in order to promote the proper application of the margin of appreciation. In my view, it has no place in the national court’s consideration of whether a measure which interferes with a Convention right is proportionate, since, as Lord Mance observed in the In re Recovery of Medical Costs case, at the domestic level, the margin of appreciation is not applicable. Indeed, in the national setting, this court, in a number of cases, has articulated an approach to examination of the proportionality of the interference where consideration of the question whether it was “manifestly without reasonable foundation” is conspicuously absent.

As Lord Reed said in Bank Mellat (No 1) [2013] UKSC 39; [2014] AC 700, pp 789-790, para 71:

This is an important statement. It emphasises that not only is the technique adopted by the national court to the question of the proportionality of a measure different from that of the Strasbourg court but the basis on which there should be reticence on the part of a national court to interfere is also different. In Strasbourg it is recognised that the court may be “less well placed than a national court to decide whether an appropriate balance has been struck”. By contrast, the national court may consider itself constrained by “national traditions and institutional culture”. One can quite see how the concept of ‘manifestly without reasonable foundation’ assists in the examination by the Strasbourg court of the proportionality of a measure. Very different considerations arise when the national court examines proportionality.

The steps in the proportionality analysis at the national level are well settled. When considering whether legislative measures which interfere with a Convention right satisfy the requirements of proportionality, “four questions generally arise”, as Lord Wilson explained in R (Aguilar Quila) v Secretary of State for the Home Department [2012] 1 AC 621, para 45. These were discussed by Lord Reed in Bank Mellat at paras 20ff:

Has the manifestly without reasonable foundation formula any part to play in the answer to be given to any of these questions? In R (SG (previously JS)) v Secretary of State for Work and Pensions [2015] UKSC 16; [2015] 1 WLR 1449, paras 209 and 210, Lady Hale addressed this issue:

I was one of those who accepted in the SG case that the manifestly without reasonable foundation test applied to all of the stages in the proportionality analysis. I was wrong to have done so. In the In re Recovery of Medical Costs case Lord Mance in para 46 had drawn a distinction between the first three stages of the proportionality assessment and the final stage:

At para 51 Lord Mance referred to the Axa case and pointed out that both Lord Hope and Lord Reed had treated the questions of ‘legitimate aim’ and whether the measure was proportionate separately. The question of whether the measure pursued a legitimate aim was to be determined on the basis that it should be considered to have done so unless the claim that it did was manifestly unreasonable. But in relation to proportionality, as Lord Mance observed, Lord Hope applied the fair balance test, citing Sporrong and Lönnroth v Sweden (1982) 5 EHRR 35 and Pressos Compania Naviera SA v Belgium (1995) 21 EHRR 301.

It is clear from this that, even in the supra-national context of Strasbourg review, a distinction is to be drawn between, on the one hand, the pursuit of a legitimate aim for the measure and, on the other, the balancing of the interests of the state against the impact which a measure interfering with a Convention right has on those affected by it. The inaptness of applying the manifestly without reasonable foundation standard when the matter of where the balance lies is being considered by a national court is all the more obvious.

I have concluded, therefore, that, certainly so far as concerns the final stage in the proportionality analysis, the manifestly without reasonable foundation standard should not be applied. Quite apart from the imperative provided by the authorities, I consider that to impose on the appellants the obligation of showing that a measure is manifestly without reasonable foundation is objectionable for two reasons: firstly, it requires proof of a negative; secondly, and more importantly, much, if not all, of the material on which a judgment as to whether there is a reasonable foundation for the measure will customarily be in the hands of the decision-maker and not readily accessible to the person who seeks to challenge the proportionality of the measure which interferes with their Convention rights. The proper test to apply in relation to the final stage of the proportionality assessment is whether the government has established that there is a reasonable foundation for its conclusion that a fair balance has been struck.

In para 67 of his judgment, Lord Wilson says that a move is afoot, as exemplified by observations made by me in the SG case, for UK courts to treat the UNCRC as part of our domestic law. So far as I am aware, the statements that I made there have not been taken up by any court and I make clear that, for the purposes of this case, I do not rely on the view that I there expressed, nor do I wish to reopen a debate on that issue. I am content to adopt Lord Wilson’s approach to the significance of UNCRC in the resolution of this case. In particular, I agree with his description (in para 68) of the elements of the concept of the “best interests” of the child in article 3.1. As Lord Wilson says, authoritative guidance is to be found in para 6 of General Comment No 14 (2013) of the UN Committee on the Rights of the Child. This was recognised by Lord Carnwath in the SG case and endorsed by this court in the Mathieson case. Lord Wilson has suggested, and I agree, that the concept has three dimensions. The most important of these, so far as concerns the present case, is the third, namely, that, whenever a decision is to be made that will affect an identified group of children, the decision-making process must include an evaluation of the possible impact of the decision on them.

That obligation, when allied to the duty of the government to reach a decision which is proportionate, according to the principles earlier discussed, means that the respondent must assess the impact on the children in a way that balances their interests against the interests of the community. It must, moreover, be satisfied that the decision strikes a fair balance and must be in a position, when challenged, to establish that it has a reasonable foundation.

The provisions in articles 26 and 27 of UNCRC, set out by Lord Wilson in para 70 of his judgment, provide an essential backdrop to that exercise. Thus, the state’s duty is to take necessary measures to ensure that children’s right to social security benefits is fully realised; to recognise children’s entitlement to an adequate standard of living; and to take measures to assist parents to implement the right to a proper standard of living, particularly with regard to nutrition, clothing and housing.

Even if these provisions are not directly binding on the government (which is the premise on which I am prepared to proceed in the present case), they are central to the judgment which the state must make in introducing social security measures which will affect the living conditions of children. Moreover, they provide an indispensable yardstick against which the proportionality of the measures under attack in this appeal falls to be examined.

For the reasons given by Lord Wilson in paras 75 and 76, the government cannot assert that the measures under attack are not directed at children. And I agree with his conclusion in para 78 that the interests of the lone parents in the present appeals are indistinguishable from the interests of their children below school age. The issue which is then starkly exposed as critical to the outcome of this appeal is the effect of article 3.1 on the proportionality of the government’s decision not to exempt from the revised benefit cap the cohorts whom the appellants represent.

I say “the effect of article 3.1 on proportionality” advisedly. Lord Wilson has carried out (in paras 81-86) a comprehensive review of the Parliamentary materials and other policy documents which disclose the background to the government’s decision and its objectives. I have nothing to add to that review. Where I part company with Lord Wilson, however, is in his concentration on the question whether the government has acted in breach of article 3 of UNCRC.

Lord Wilson concludes “by a narrow margin” that the government did not breach article 3.1 by its refusal to amend the 2006 Regulations so as to exempt the appellant cohorts from the revised cap. As I understand his judgment, largely on that account, he considers that the appeal must be dismissed. In a telling sentence in para 87 he says:

I do not agree that the questionability of the government’s decision or its avowed lack of balance should not be addressed by this court. Conclusions on those matters will not - at least, not necessarily - be determinative of the appeal. But, inasmuch as they sound on the question of the proportionality of the government’s decision, they are matters to be taken into account. I will return to this theme in paras 188-190 of this judgment.

In the meantime, it is important to deal with the significance to be attached to a finding that the government has not acted in breach of article 3 of UNCRC. One may begin by recognising that, of course, if the government was found to have acted in breach of that provision, this would go a long way towards showing that the decision not to exempt the appellant cohorts from the revised cap was disproportionate, if indeed it would not be conclusive on that issue. But a finding that no breach of article 3 arose does not establish the converse.

But I query the premise (which I believe to be implicit in Lord Wilson’s judgment) that the question of whether the government was in breach of the article is pivotal to the issue of proportionality. UNCRC contains a number of enjoinders to those countries which subscribe to it. Some of these are expressed in imperative terms. The duty of the state is to keep faith with the spirit of the Convention. Whether it has discharged that duty is not to be answered solely on whether it can be said to be in technical breach of its terms. The proportionality of a government measure which has an impact on the best interests of children is not to be judged by a mechanistic approach to the question whether there has been technical compliance with article 3. It must be assessed on the basis of whether, given the injunctions in UNCRC, the government’s decision, taking into account where the best interests of children lie, represents a balanced reaction to those interests and the aims which a particular measure seeks to achieve. I should say, however, if the proportionality of the government’s decision not to exempt the appellants from the benefit cap depended on whether there was a failure to comply with it, I would have held that the government was in breach of article 3. I will discuss the reasons for that conclusion later in this judgment.

Article 3 (and articles 26 and 27) provide a context as well as a backdrop to the government’s decision as to those who should be covered by the cap. That decision is not insulated from challenge on proportionality grounds by the government’s claim that it took representations into account, nor even that it carried out an evaluation of their weight and persuasiveness. The government must show that it reached a balanced conclusion, taking into account the impact which the refusal to exempt the cohorts whom DA and DS represent has had upon them, when weighed against the interests of society which the conclusion is said to protect.

The impact of the decision not to exempt the DA and DS groups is well described in the submissions of Gingerbread made to the Public Bill Committee of the House of Commons in September 2015, referred to in para 83 of Lord Wilson’s judgment. No real answer to the criticisms of the scheme has been provided by the government. Its principal defence is its reliance on the DHP scheme. The shortcomings of that scheme have been vividly described in paras 30 and 31 of Lord Wilson’s judgment. Quite apart from the myriad of difficulties to which he there refers, the fundamental point to be made is that DHPs are not tailored to deal with the spectrum of difficulties which the appellants face, merely one aspect of them: housing costs. They do nothing to alleviate problems with childcare costs and complications in obtaining childcare, even if it could be afforded. And, of course, there is, as Lord Wilson pointed out in para 31, scant, indeed, virtually no, information as to the extent by which the difficulties encountered by the DA and DS cohorts are mitigated by DHPs. There is simply no warrant for the claim that refusal to extend exemption from the cap to the DA and DS cohorts will improve the fairness of the social security system or increase public confidence in its fairness. That sweeping statement partakes of a declamation for which no tangible evidence is proffered. To the contrary, a proper understanding of the impact on those whom the appellants represent, so far from increasing public confidence in the social security system, is likely to lead any right-thinking person to the opposite conclusion.

The other two professed aims of government, to incentivise parents in a non-working family to obtain work and to achieve fiscal savings have been decisively refuted by the evidence. One can only incentivise parents to obtain work if that is a viable option. The evidence in this case overwhelmingly shows that in most cases in the DA and DS cohorts, this is simply not feasible. In particular, lone parents are placed in an impossible dilemma. If they go out to work, they must find the resources for childcare. Those in the DA and DS groups will routinely find it impossible to obtain employment which will remunerate them sufficiently to make this a sensible choice. They also face the difficulty of obtaining suitable childcare, irrespective of whether they can afford it.

As to the fiscal savings that might be achieved, Lord Wilson has dealt summarily and conclusively with that argument in para 32 of his judgment. I agree entirely with what he has had to say there and need not repeat it.

The enjoinder in article 3.1 of UNCRC that, in all actions concerning children undertaken by administrative authorities or legislative bodies, the best interests of the child shall be a primary consideration sets the scene for an examination of whether the failure to exempt the DA and DS cohorts from the cap is a proportionate interference with their ECHR rights. It is to be noted that the best interests of the child must be a primary consideration. Where those interests conflict with other considerations, although they will not inevitably prevail, their primary status must be respected. Ephemeral aspirations, however high-sounding or apparently noble, will not suffice to displace them.

The entitlement of children, enshrined in articles 26 and 27 of UNCRC, to have the state take necessary measures to ensure that their right to social security benefits is fully realised; and that this comprises an adequate standard of living; and that measures must be taken to assist parents to implement that right all contribute to the importance that UNCRC places on the welfare of children. Where measures are adopted by a state which have a demonstrable adverse effect on children, the hurdle faced by government in showing that these factors have been properly taken into account is correspondingly heightened.

Government, if it is to adhere to its obligations under UNCRC, must have a clear-sighted understanding of the impact on children that a proposed measure curtailing their entitlement to social security benefits will have. It must also carry out a defensible weighing of their interests against the objectives which it proposes will be achieved by a curtailment of the rights. The preponderance of evidence in this case strongly supports the conclusion that this is not the way in which the respondent approached the decision under challenge in this appeal. True it is that it considered the representations made. But I do not conclude that a proper weighing of the particular interests of the DA and DS cohorts against what was likely to be achieved in their case was carried out.

It is not enough that notice was taken of the various submissions made, or that the amendments proposed to the scheme came to the attention of the government. There must be a frank and objective assessment of whether depriving these particular individuals of the benefit of exemption from the cap would conduce in a material way to the realisation of the avowed aims of the scheme. For the reasons that I have given, I do not consider that such an exercise was undertaken. Further, I believe that, if it had been, the case for the inclusion of the DA and DS cohorts in the exemption should have been found to be irresistible.

As I have said, I do not consider that breach of article 3 of UNCRC is an essential prerequisite to a finding that there has been a disproportionate interference with the appellants’ ECHR rights. But, if such a breach required to be found, I would have concluded that it was present. The evidence in this case unmistakably points to the inference that, while the impact on children’s rights was considered, it was not given a primacy of importance which article 3 requires. Had it been, the conclusion that the exemption should not be extended to the DA and DS cohorts would not have been reached.

Lord Wilson has amply demonstrated that government and Parliament were alive to the state’s obligation under article 3 of UNCRC. Where, regretfully, I must disagree with him is on his conclusion that taking into account the representations made amounted to a discharge of that obligation.

I would allow the appeal and make a declaration that the regulations represent an unjustifiable interference with the appellants’ article 1, Protocol 1 and article 8 rights, taken in combination with article 14 of ECHR.