England and Wales High Court (Administrative Court) Decisions

IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
ADMINISTRATIVE COURT

		Royal Courts of Justice Strand, London, WC2A 2LL
		17/08/2018

The Hon Mr Justice Kerr:

Summary

1. The defendant Secretary of State has placed a limit on (capped) the amount of certain payments made under a discretionary assistance scheme established under statutory powers to help people with disabilities to work. The claimant is affected by this and says the cap is unlawful because of a breach of the public sector equality duty and because the cap is indirect discrimination on the ground of disability.

Facts

2. The Employment and Training Act 1973 (the 1973 Act), section 2(1), requires the Secretary of State to make such arrangements as he or she considers appropriate for the purpose of assisting persons to select, train for, obtain and retain employment suitable for their ages and capacities or of assisting persons to obtain suitable employees.
3. This may include (section 2(2)(b) and (d)) arrangements to encourage increases in employment and training opportunities for, among others, disabled persons; and may include making payments to persons who provide or use facilities provided for as part of the arrangements or to other persons, as specified in the arrangements.
4. The Secretary of State exercised that function in 1994 by setting up the Access to Work Scheme (the Scheme). It provides for funding contributions to increase employment opportunities for disabled people, according to fixed criteria. It operates within a finite budget set by the Treasury, but is demand led, in the sense that payments are made according to the Scheme criteria and are not limited by any ceiling on the total annual amount of payments.
5. In December 2010, the then government commissioned Ms Liz Sayce OBE to conduct an independent review into specialist disability programmes including the Scheme and its impact for disabled people. The terms of reference required recommendations on how to achieve better value for money but the recommendations had to be "deliverable within the existing funding envelope…".
6. The claimant, Mr Buxton, is a user of the Scheme and is deaf. Current terminology on those with impaired or no hearing is explained in a footnote at the start of the claimant's skeleton argument. The claimant is "Deaf", i.e. his first language is British Sign Language (BSL) and he "identifies as culturally Deaf and part of the Deaf community".
7. Here, I use the word "deaf" to mean, unless the context requires otherwise, a person who is either "Deaf" in that sense, or is "D/deaf", a term that includes in addition those who are "hard of hearing" but "do not identify as culturally Deaf and/or [who] have English as their first language and use lip-reading and hearing aids to communicate".
8. Mr Buxton began working for the charity Scope in 2006. He received funding under the Scheme. He was working there when, in 2010, the then government commissioned an independent review by Ms Sayce of its specialist disability programmes, including the Scheme. Her report (the Sayce report) was called Getting in, staying in and getting on; disability employment support fit for the future. It was published in June 2011.
9. I was taken to passages in the Sayce report. It recommended increased funding for the Scheme and increased coverage for more disabled people. It suggested the Treasury would achieve a net return of £1.48 for every pound spent. It recommended increased publicity to encourage more applicants. Value for money would be achieved, broadly, by investment in disabled persons helping them to become more productive.
10. The Sayce report stated that the proportion of the budget spent on existing users was increasing year on year. Use of the Scheme was disproportionately low among people with mental health and learning difficulties. Statistics were included, breaking down the then 37,290 users by type of impairment. Persons with back and neck problems, visual and hearing impairments were at or near the top of the list by number of users.
11. In 2011, Mr Buxton moved from his job at Scope to a new post at the British Deaf Association. He continued to receive support from the Scheme, at a decreased level because he was working mainly with other users of BSL and therefore required a lower level of support. New guidance on use of the Scheme was introduced at the time the Sayce report was published.
12. Public consultation on disability employment support took place in the latter part of 2011. In March 2012, the government announced a command paper on strategy for disability employment programmes and confirmed that it accepted the recommendations in the Sayce report, subject to setting up an expert panel to hold further discussions with interested parties and consider how the recommendations should be implemented.
13. In 2012 or 2013, the government amended its guidance on the Scheme to introduce what became known as the "30 hour rule": if a full time support worker were required, e.g. for 30 hours or more each week, that should normally be funded on an annual salary basis rather than a freelance hourly basis. This did not work well for deaf users of the Scheme.
14. The reason was that many BSL interpreters, in high demand, preferred to work by the hour and few were willing to work on a salary basis. In June 2014, during the course of the expert panel's discussions and deliberations, the then minister announced suspension of the 30 hour rule for new Scheme claimants. The minister also announced an internal departmental review of the Scheme.
15. At about the same time, the House of Commons Work and Pensions Select Committee (the Select Committee) announced that it too would be enquiring into the Scheme. Meetings and events were held and there were contributions from interested parties including charities working for people with disabilities and providers of BSL interpretation services.
16. The Select Committee reported in December 2014. The report emphasised that attracting new entrants with only a marginally increased budget created a risk of "bearing down on the awards of people who happen to have relatively high cost needs". The report recommended increased funding to avoid this and discussed certain technical options for persuading the Treasury to agree.
17. In the report, the Select Committee noted that the minister did not support the suggested £1.48 net return for every pound spent and that he had asked officials to "undertake some analytical work" on the cost effectiveness of the Scheme, which would be necessary to persuade the Treasury to agree to increased funding.
18. The committee welcomed the Secretary of State's intention to increase the number of recipients of Scheme funding. It considered that priority should be given to "supporting people with mental health problems, and other more hidden intellectual, cognitive and behavioural impairments, and learning disabilities, to gain and continue in employment".
19. In relation to the suspended 30 hour rule and the effect on deaf people in need of BSL interpreters, the Select Committee recommended that the Department for Work and Pensions (the department) should consult the BSL interpreting profession to establish agreed maximum hourly rates and should only apply the 30 hour rule, if at all, in cases where a salaried interpreter was known to be actually available.
20. In the wake of the Select Committee report and the Sayce report, Mr Buxton and others took part in three meetings with government, at which he represented the British Deaf Association which, in turn, represented the UK Council on Deafness in the discussions. These took place from December 2014 to February 2015. In March 2015, the Scheme guidance was amended to exclude BSL interpretation from the scope of the 30 hour rule.
21. In advance of the government's response to these developments, a draft equality analysis was prepared for consideration before making the decision on how to respond. I have the subsequent final version, published in May 2015 and featuring, I am told, minor differences from the draft version considered by the Secretary of State before announcing the decision. It is not necessary to focus on the differences between the draft and final versions.
22. The equality analysis of 2015 (the EA 2015) began by setting out the content of the public sector equality duty under section 149 of the Equality Act 2010 (the 2010 Act). The nature of the Scheme was then outlined. It was emphasised that the Scheme is not intended to replace reasonable adjustments the employer is obliged to make, but involves provision over and above that; and that provision should meet the minimum needs of the "customer", i.e. the disabled person, in the most cost effective way.
23. The 2015 EA went on to note the challenge of running a demand led Scheme on a finite budget while increasing the number of funding recipients. A balance had to be struck between "meeting customer need" and "achieving value for money". Among ways of achieving the latter aim would be the setting of caps on "high-value awards", with "transitional protection".
24. Capping was then considered in detail. To achieve the right balance between the need to support as many users as possible, and what it is reasonable to offer individual users, the option of setting a "cap on the maximum value of support per user" was considered. While such a cap might be seen as "arbitrary", setting a cap at "a multiple of average salary" would reflect the point that "many high-value awards purchase the wages of support workers".
25. Potentially available savings were then tabulated, assuming the setting of a cap at three differing rates. This showed the amount of money that would be freed up by setting the cap at each of the three levels and the approximate number of additional users who might receive an award, averaged at the 2013-14 average level of award, namely £3,045.
26. The number of users affected by a cap was then forecast in a table. If the cap was set at the highest of the three proposed levels, namely double average salary, or £54.4k, the answer was 79 people, or 0.2 per cent of users as at 2014-15; 94 per cent of them deaf. The amount saved would be 1 per cent of the overall budget of about £124 million.
27. At the other end of the spectrum, if the cap were set at full time average salary, or £27.2k, 540 people would be affected, amounting to 2 per cent of users, 67 per cent of them deaf; making a saving of £7.92 million, or 8 per cent of the overall budget. The observation was made that wherever on that spectrum the amount of the cap were set, the majority of those affected would be deaf users.
28. The 2015 EA went on to consider the "[e]quality impact". An unknown feature discussed was whether employers would cover the shortfall or whether there might be damage to relations between disabled and non-disabled people, and whether employers might be put off employing disabled people. Lacking "the evidence to be able to quantify either of these risks", the department would "take steps to monitor the impact and … consider if further flexibilities are required should an adverse impact be found".
29. Next, it was noted that deaf users then receiving high value awards may be "especially affected by capping". Again, the evidence needed to quantify the impact was lacking, though the tabulated information did provide a "worst case scenario". The department would "continue to monitor the impact" and consider further changes should it detect "an adverse impact on disabled people in general, or upon a particular group of disabled people… ."
30. What are called mitigation measures, to soften the blow to those who may be adversely affected, were then considered. There were three: first, a possible change in employer behaviour; second, increased access to "technology solutions and training"; and third, transitional cushioning, giving users time to adapt to the cap before its arrival. A notice period of three years for existing users and of six months for new users was proposed.
31. While it was "difficult to quantify what effect these mitigating factors … would have on the equality impact of capping", the department would "continue to monitor this if capping is implemented and will consider further flexibilities if an adverse impact upon equality emerges". It was noted that to increase the numbers of disabled people using the Scheme is to support the rights of disabled people to work equally with others. A cap achieves this, while maintaining "significant levels" of funding for existing affected users.
32. The EA 2015 included a later, separate section on the position relating to BSL interpreters. Responding to the Select Committee's proposal for agreed hourly rates for BSL interpreters, the EA 2015 proposed "central contracting" through a commercial framework agreement. This, it later transpired, was not successfully negotiated. The 30 hour rule should, it was suggested, be discontinued altogether; its particular impact on deaf people was noted.
33. On 12 March 2015, having considered the draft version of that assessment, the Minister for Disabled People announced the government's decision in the House of Commons. From October 2015, awards would be capped at one and a half times average annual salary, setting a limit of £40,800 per person per year, uprated annually. Those receiving more than £40,800 would not be subject to the cap until April 2018.
34. The three year notice period would, said the minister, help them and their employers adjust; specialist teams from the department would help with advice on reasonable adjustments and greater use of technology. Mention was also made of discussions to develop a framework agreement for translation services including BSL interpreters which, it was hoped, would lead to transparent rates from the summer of 2015. The 30 hour rule was abolished.
35. The final version of the EA 2015 was published in May 2015. The proposed framework agreement with agreed hourly maximum rates for BSL interpreting services did not come into being in the summer of 2015, and has not done so since. Then in September 2015, just before the cap was to start for new Scheme users, the government responded to the Select Committee's report.
36. The cap, said the government, would affect 200 existing users whose awards would be protected until April 2018. In response to the Select Committee's invitation to undertake research to support a request for additional funds, the government said that officials "have undertaken research to explore the potential costs and benefits of Access to Work expenditure including the impacts on social security and income tax returns".
37. However, no estimated or forecast figures had emerged from this exercise, mainly because there was no available "comparison group against which to assess the difference the programme makes to someone's likelihood of being in work". The government would continue to "build the evidence base with stakeholders' input to understand the value Access to Work adds".
38. The cap then came into force from October 2015. The department wrote to Scheme funding recipients about the changes. The British Deaf Association met the minister to discuss the changes. Evidence on the market for BSL interpretation services was called for. Mr Buxton was involved in these activities in his capacity as a representative of the British Deaf Association.
39. During 2016, Mr Buxton applied successfully for the job of chief executive officer of Action for Disability. But he did not immediately take up the post. In November 2016, he emailed the department raising concerns about his support needs in his new job and the impact of the cap, which would apply to his reassessed award once he took up the post. Mr Buxton's request was for annual BSL support of over £100,000, far in excess of the cap.
40. The department wrote to him on 12 April 2017 awarding him £37,847.75 for the period from 8 May 2017, the start date for his new job, until 31 March 2018, the end date for transitional protection. The cap then stood at £42,100 per annum, but Mr Buxton's award was protected at the October 2015 level of his award. Thus, the £37,847.75 represented his annual award as at October 2015, prorated for the part year period from 8 May 2017 to 31 March 2018.
41. Mr Buxton was not happy with the award. He started his new job as CEO of Action for Disability on 8 May 2017. He experienced difficulty making himself understood by his staff because, he explained in his witness statement, he did not have enough access to BSL interpreter services. The UK Council on Deafness was also, at the time, lobbying to get the cap overturned. Mr Buxton was concerned both on his own behalf and for deaf people generally.
42. The department's witness, Mr Stuart Edwards, has taken issue in his evidence with Mr Buxton's claimed assessment of his own support needs; he maintains that Mr Buxton had available to him support worth over £62,000, taking account of adjustments voluntarily made by his employer. I need not resolve this satellite dispute which is relevant to neither ground of challenge. There is no rationality challenge to the assessment of Mr Buxton's personal award.
43. In July 2017, the department published a "market review" of BSL provision. The main problem was that demand far outstripped supply. The review found that most good and reputable full time BSL interpreters would work for around £250 per full day in London and the south, representing an hourly rate of about £35 for a seven hour day; whereas agencies would often charge £300 to £500 for a full day's work.
44. In mid-July 2017, Mr Buxton applied for reconsideration of his award. He argued in detail against the cap in principle, as he considered it unfair, and against the amount of his own award, arguing that his needs had changed. The response letter of 27 July 2017 explained the reasoning in more detail but did not alter the decision. The transitional protection up to 31 March 2018 apparently did not extend, at least in Mr Buxton's case, to increasing an uncapped award due to increased need, for the period down to 31 March 2018.
45. The matter then moved in the direction of litigation, with pre-action protocol correspondence exchanged. The claim was issued on a protective basis, on 26 October 2017, challenging the decision of 27 July 2017 to cap Mr Buxton's award and seeking a declaration that the staff guidance giving effect to the cap is unlawful. Damages were also claimed.
46. The government and MPs continued to consider the matter after the claim was issued. Transitionally protected users were invited to have discussions with specialist staff from the department, to discuss managed personal budgets and workplace assessments. Officials of the department met representatives of charities active in the sector, including the UK Council on Deafness.
47. In January 2018, Cockerill J granted permission for the claim to proceed, limited to the two grounds now before me. She rejected as unarguable the proposition that the Scheme, being discretionary in nature, was a social security scheme for the purposes of article 1 of the first protocol to the European Convention, and article 14 thereof, and she rejected as unarguable a proposed irrationality ground of challenge.
48. In February and March 2018, the equality analysis in EA 2015 was updated by a further document (the EA 2018). I have the version that was considered by the minister, as well as the later published version. The earlier version looked at the numbers affected by the cap and those afforded transitional protection up to 31 March 2018. It was noted that there was a division of opinion between those opposed to any cap, in principle, and those not so opposed.
49. Reference was made to the twin aims of increasing the contribution of employers to provision of support, and liberating resources to enable new entrants to join the Scheme. Five options were considered, one of which was to leave the 2015 decision unchanged. The other four involved changing either the level of the cap, or the duration of transitional protection, or both, or introducing further "supporting mitigations".
50. The impact of the cap on those affected, and the impact on those whose transitional protection was about to end, was considered with some statistics and tabulated information. About 42 per cent, or about 80 of the 200 people entitled to transitional protection, had reduced their spend to below the April 2018 cap of £43,100. This had saved £2.4 million, which had become available to support new entrants to the Scheme.
51. There was clear recognition that disability type was the main feature distinguishing capped and uncapped users of the Scheme. Deaf people were disproportionately affected by the cap, along with, to a lesser extent, the visually impaired. This was because their support needs, principally for BSL interpreters, were the most expensive. Concerns were noted that the cap could prejudice their position at work, isolating them and acting as a "glass ceiling".
52. The response to these arguments in the EA 2018 could be characterised by the French proverb which translates as "you cannot make an omelette without breaking eggs". The Scheme was not intended to eliminate all difficulties experienced by people with every type of disability. Distribution of the finite budget as between users with different disability types had to balance meeting individual needs against broadening the range of those receiving support.
53. The authors of the EA 2018 noted, however, that evidence of a positive response from employers was scant and disappointing; "further behaviour change is needed because employers are not sufficiently discharging their statutory reasonable adjustment duties". They recognised the adverse impact on the 120 or so individuals affected by the cap, whose needs remained in excess of it. The conclusion was to reject the status quo:

"… whilst we maintain that the current cap in principle is necessary and proportionate in light of the points made above in its favour, we also consider the potential impact of varying the cap or [transitional protection] level/duration."

54. After discussion of the likely financial impact of the other options, the authors favoured "Option D", which involved no change to the transitional protection ceasing from 31 March 2018, but raising the cap from one and a half times to twice average annual salary, despite the resulting reduction in savings of about £975k. This would mean, they estimated, that over 80 per cent of the original transitionally protected group would be brought within the level of the cap.
55. They also estimated that assuming a working year of 230 days, an individual award would purchase support of just under £250 per day, which closely corresponds to the figure of £260 for a full day of BSL interpretation, according to the suggested rates published by the National Union of British Sign Language Interpreters (NUBSLI), for London and the south east; the suggested daily rate for the north east being lower, at £210.
56. The EA 2018 ended by saying that the department would continue to monitor the impact of the cap, in particular by recording changes to employers, jobs, roles, hours worked and seniority. There was no analysis of any research done into the cost effectiveness of increasing funding for the Scheme and no evidence that any progress has been made on that front.
57. After considering the EA 2018 as it then stood, the Secretary of State told the House on 20 March 2018 that the average spend among the remaining transitionally protected Scheme users had decreased from about £57,000 each to about £45,000 each; that this suggested it was achieving its "intended incentive effects on individuals and employers to make best use of funding as well as freeing over £2 million per year, to support growing numbers …".
58. The minister declared herself "persuaded that the principle of the cap is sound, balancing the need to provide support to the largest number of people, and at a significant level for some, with the need to make the best use of public funds". She announced that the cap would rise to £57,200 per year, double average annual earnings, from its then current level of £43,100.
59. There would be further measures to help those affected: "extra support to customers with high-value awards via automatic workplace assessments promoting available technology and reasonable adjustments and voluntary cost-share from employers" and, among other things, "continued monitoring of the impacts [of] the cap" and "discretion in exceptional cases of multiple disability" to apply different rules.
60. On 24 March 2018, Mr Buxton made a request from the department for funding support of £67,200 for the year from 1 April 2018 to 31 March 2019. The response from the department was to award him £57,200, the amount of the cap which was £10,000 less than he had requested. The decision maker recognised that he needed £280 per (working) day for a BSL interpreter.
61. In May 2018, the final version of the EA 2018 was published. There were differences from the earlier version the minister had considered. Most notably, there was an estimate of "behavioural impact" of setting the cap at twice average annual salary, which though "difficult to quantify" was thought to be "around £3.4mpa" (millions of pounds per year). This figure did not feature in the version dating from February or March 2018.
62. Mr Buxton applied, on 31 May 2018, to amend his application so as to challenge, on the same grounds as before, the decision of the minister announced on 20 March 2018. The Secretary of State did not agree to the amendment, but accepted before me that I should grant or refuse the application to amend, and permission to bring the claim as amended, according to whether I consider the amended case to be arguable.

First Ground: Public Sector Equality Duty

63. It is said there was a breach of the duty enacted by section 149 of the 2010 Act. The issue is whether the Secretary of State had "due regard" to the three legislative goals set out in section 149(1)(a), (b) and (c): the need to eliminate discrimination, harassment, victimisation and other conduct prohibited by or under the 2010 Act; to advance equality of opportunity between those with a relevant protected characteristic and those without it; and to foster good relations between those with, and those without, a protected characteristic.
64. Advancing equality of opportunity under section 149(1)(b) involves (see section 149(3)(c)) having due regard, in particular, to the need to remove or minimise disadvantages suffered by persons who share a protected characteristic, the need to take steps to meet the needs of persons who share a relevant protected characteristic and those who do not; and encouraging persons who share a relevant protected characteristic to participate in public life or other activities in which participation by such persons is disproportionately low.
65. No issue of law divides the parties. The issue is one of fact; did the Secretary of State have the necessary "due regard" to the matters set out in the section? The nature of the "due regard" obligation has been articulated by the higher courts in many well known cases. Since there was no difference between the parties on the law and this is now well travelled ground, it is unnecessary to repeat the propositions derived from those cases.
66. For Mr Buxton, it is submitted that when deciding to revise the cap from 1.5 times average annual salary to double average annual salary, the Secretary of State failed to do enough to ascertain and assess the true impact of the revised cap on deaf people, just as her predecessor in 2015 had failed to do so when imposing the initial cap of 1.5 times average annual salary. In summary, he submits that the enquiries were too superficial, the statistics are too bald and the impact on human lives that underlies them was insufficiently explored.
67. The first part of the argument asserts a failure to do the research needed to make the assessment of the effect on the lives of those affected. Mr Buxton complains that there was not enough information about employer behaviour since the 2015 cap; having acknowledged in 2015 that increased assistance from employers was needed and monitoring would be undertaken, that was not done. Nor was there an adequate assessment of the effect of the revised cap on the deaf people most affected by it.
68. In more detail, Mr Buxton argues that the sub-groups of those affected should have been identified and the effect on each analysed separately: those without transitional protection or whose transitional protection had ended, including Mr Buxton; those entitled to transitional protection who had managed to reduce their spend; those entitled to transitional protection who had not managed to reduce their spend; and those who had not applied at all, or had applied for a lower spend than they would have done but for the 2015 cap.
69. Ms Hannett, for Mr Buxton, engaged in a long and detailed written critique of the equality analysis in the EA 2015 and EA 2018. It amounted to an argument that the impact of the cap on the various categories of Scheme users was not sufficiently assessed. The questions were not sufficiently searching. Had employers made up the shortfall in provision since the 2015 cap arrived in October that year? What did it mean in practice to reduce one's spend? What were those who did so going without?
70. These kinds of questions were, Ms Hannett submitted, not answered because they were not asked by the Secretary of State of himself in 2015, nor of herself in 2018 when the updating EA 2018 was produced. For example, the 2018 EA says that in the case of the approximately 40 people entitled to transitional protection who had not managed to reduce their spend, the revised cap would mean that 90 per cent of their assessed needs were met. What about the loss of the other 10 per cent? The EA 2018, it is said, does not provide a "qualitative analysis" of what losing that 10 per cent would mean in practice.
71. Further, the expectation in 2015 that employers would step up their provision of reasonable adjustments to perform properly their legal duty to do so, had not been realised by March 2018. The provision of such adjustments, for example in the form of auxiliary aids, had not been measured. This approach is in any case flawed, submitted Ms Hannett, because the Scheme is supposed to provide support for adjustments the employer is not required to provide; it is not supposed to duplicate adjustments the employer is required to provide.
72. Mr Buxton also charges the minister with a failure to heed what is said in section 149(3)(c) of the 2010 Act. The successive secretaries of state had both given higher priority to increasing the number of participants in the Scheme than to maintaining the standard of provision for those already in it. Yet, said Ms Hannett, participation is not just about numbers; it is also about the level at which participants can take part.
73. The two EAs, she said, had failed to grapple with "the fact that those affected by the Cap are likely to be D/deaf people in professional, managerial or leadership roles …". It was said that the Secretary of State dismissed as irrelevant concerns that the cap would act as a glass ceiling, particularly for deaf people. The width of the enquiry required to comply with the public sector equality duty could not be cut down by ignoring the capacity in which Scheme participants are employed, as well as the fact of being employed at all.
74. It was argued that this approach overlooked the need stated in section 149(3)(a) of the 2010 Act to remove or minimise disadvantages suffered by persons with a protected characteristic that are connected to that characteristic; and the need articulated in section 149(3)(b) to take steps to meet the needs of persons who share one of the characteristics that are different from those who do not share it. Finally, it was said that the EA 2018 did not address "the duty to [have due regard to the need to] foster good relations between disabled and non-disabled people": section 149(1)(c) of the 2010 Act.
75. For the Secretary of State, Ms Leventhal emphasised that the purpose of the exercise was to improve the position of disabled people in the workplace, albeit subject to financial constraints. The Scheme operates to contribute to the needs of its disabled users; it is not designed or required to meet all their needs. The decisions, first to impose a cap and then to raise it, were properly concerned to address equitable distribution of available funds as between people with different types of disability, while achieving value for money.
76. Mr Buxton's attack on the equality analysis amounted to no more than "forensic criticisms" of the EA 2015 and EA 2018 and disagreement with the result of the exercise, rather than establishing any failure to have due regard to the legislative goals set out in section 149 of the 2010 Act. There may be disappointment that the cap was not raised further or abolished in 2018; but it is unreal to say that raising the cap, which impacted most on deaf people, involved a failure adequately to address the needs of those same deaf people.
77. Ms Leventhal pointed out that the two EAs had to be seen, not in a vacuum but in the context of an evolving debate going back to the Sayce review in 2010-11, the Select Committee report, and numerous meetings and discussions with campaigning organisations and representatives of disabled people. It was for the minister, not the court, to determine what weight should be given to equality considerations when performing the "due regard" duty, and to decide how to allocate available funds as between people with different types of disability and as between existing Scheme users and potential new ones.
78. I accept the Secretary of State's submissions and I do not think those of Mr Buxton even raise an arguable case that it was unlawful to cap his July 2018 award because of a prior breach of the public sector equality duty. A decision on how to allocate resources as between different categories of persons sharing the same protected characteristic (albeit in different ways) was always likely to be an ambitious target for an allegation of breach of the equality duty.
79. You cannot carry out such an exercise at all without having some regard (whether or not short of "due" regard) to the needs of the people affected by it who share that characteristic. The allocation of funds to disabled people with varying disabilities is likely to amount to deciding how best to help disabled people fare better with and in their work than before, by comparison with each other and with their non-disabled counterparts in the workplace.
80. This exercise was therefore, from its inception, steeped in advancement of equality of opportunity and fostering of good relations between those with and those without the protected characteristic of disability. To state the obvious, there is much in the 2015 and 2018 EAs, and the ministerial decisions that followed each of them, to encourage employers to take on disabled workers, particularly those intended to gain access to the Scheme for the first time as a result of the redistribution made possible by the cap.
81. To state the obvious again, the allocation decisions criticised by Mr Buxton in this case took place against a background of finite funding. There would be no more money available unless the Treasury could be persuaded by empirical evidence so far lacking, that increased funding would pay for itself by increased tax returns or reduced social security payments. The minister was therefore in the position of deciding whether, and if so how, to give with one hand and take with the other. Nothing in the equality legislation prevents that.
82. The exercise was conceptually simple: how best to strike the balance between expansion of the Scheme for new users, on the one hand, and maintaining individual high cost provision, on the other. It was, as Ms Leventhal rightly said, a matter of judgment for the minister how best to strike that balance, first in 2015 and then again in 2018. That leaves the issue whether the regard had to the equality goals in section 149 of the 2010 Act was "due" regard or fell short of it.
83. I do not think it is arguable that the research and assessment process was too superficial to amount to "due" regard. It is not the law that to comply with the equality duty, all possible research must always be carried out and assessed before a decision is made. The level of scrutiny of the proposals under consideration here was sufficient to comply with the duty, which was in the mind of the minister on both occasions; cf. the helpful discussion of the duty and recent authorities, in the judgment of Laing J at [34]-[41] in R (DAT and BNM) v. West Berkshire Council [2016] EWHC 1876 (Admin).
84. I also reject the submission that the minister wrongly took account of reasonable adjustments which employers are not obliged to make but might be willing to make; or, conversely, wrongly failed to take into account reasonable adjustments which employers are obliged to make but have not made or may or may not be willing to make in the future. This part of the argument treats the obligation, or lack of it, to provide reasonable adjustments, as a black and white issue in each case.
85. The reality is very different; there is usually room for debate about whether a desired adjustment is reasonable and obligatory for the employer, as many employment tribunal decisions demonstrate. There is no want of due regard for the legislative aims where a decision maker considers, without exhaustive research, the extent to which employers are providing such adjustments, or may or may not provide them in the future, or the extent to which their willingness to do so without compulsion has changed, or may change.
86. Nor does any arguable breach of the duty arise in this case from striking the balance, referred to above, in such a way as to favour those who have yet to gain access to the Scheme at all, to facilitate their access to jobs they do not yet have, over those already receiving high cost awards under it, to facilitate their rise to the higher echelons of the businesses and professions in which they are already employed.

For those brief reasons, the first ground of challenge is not arguable. I agree with Ms Leventhal's description of the criticisms levelled by Mr Buxton as merely forensic. I therefore do not grant permission to amend the claim to enable Mr Buxton to advance it and, if the amendment had been allowed, I would not have granted permission for the amended challenge to proceed on the public sector equality duty ground.

Second Ground: Indirect Discrimination

87. Secondly, it is said that the cap on Mr Buxton's award in July 2017 is unlawful because the cap is indirectly discriminatory, contrary to section 19 of the 2010 Act. The cap, it is submitted, is a service provided in the exercise of a public function, therefore engaging the prohibition against indirect discrimination in section 19.
88. Ms Hannett submitted that the cap is a provision, criterion or practice (PCP). It applies to recipients of Scheme funding who do not share the relevant protected characteristic, deafness, as it applies to those who do possess that characteristic; but it puts the latter at a particular disadvantage (often called group disadvantage) when compared with recipients of Scheme funding who are disabled in other ways but not deaf. The cap also puts Mr Buxton personally at that disadvantage (individual disadvantage).
89. Ms Hannett submitted that the Secretary of State, on whom the burden lies, cannot show that the cap is a proportionate means of achieving a legitimate aim. It is the cap that must be justified, not its effect. The legitimate aim, or objective, must be sufficiently important to justify limiting a fundamental right (i.e. the right to equal treatment); the measure adopted must be rationally connected to the objective; and the means chosen must be no more than is necessary to accomplish the objective (R (Elias) v. Secretary of State for Defence [2006] 1 WLR 3213 per Mummery LJ at [164]).
90. It is agreed that section 19 is engaged. Either the Secretary of State is a "service-provider", "concerned with the provision of a service to the public or a section of the public" within section 29(1), read together with section 31(3) and (4); or, if that is wrong, she is exercising "a public function that is not the provision of a service to the public or a section of the public", within section 29(6). Either way, she must not discriminate: see section 29(2) and (6).
91. The Secretary of State now accepts that the cap is a PCP; that, while it applies alike to deaf people and people with other types of disability, it subjects the former to a group disadvantage as compared with the latter; and that it puts Mr Buxton as an individual to the same disadvantage. She accepts that the test of justification is that relied on by Mr Buxton, derived from Mummery LJ's judgment in the Elias case, in turn drawing on earlier case law; and that the burden is on her to establish that the cap, and not its effect, is justified.
92. The issue, therefore, is whether the Secretary of State can show, applying those tests, that the cap is justified in point of fact. Mr Buxton says she cannot. In truth, Ms Hannett argues, the aims relied on by the Secretary of State are "premised on costs alone" and, as such, not legitimate. Ms Hannett submits that none of the four aims advanced by the Secretary of State is legitimate.
93. The first professed aim is to broaden access to the Scheme. There is no rational connection between imposing a cap and achieving that aim, Ms Hannett submits. The second aim is to induce employers to make reasonable adjustments more readily. This, it is argued, is merely an attempt to offload cost onto employers or, failing that, onto employees themselves. The third stated aim of obtaining value for money is merely the aim of saving money; while in the case of the fourth, to ensure fairness in the distribution of the Scheme's funds, there is no rational connection between the cap and that aim.
94. As to proportionality, Ms Hannett makes the following main points. The uncontradicted, albeit not agreed, evidence is that Scheme pays for itself by yielding £1.48 for every pound spent. Therefore the fiscal aims of the Secretary of State count for little. The savings the cap is supposed to generate are likely to be illusory because of the propensity of those subject to the cap to increase their awards to the level of the cap, according to the 2018 EA.
95. Further, Ms Hannett submitted that the Secretary of State had failed to show there is no less intrusive way of achieving her aims, in a manner that does not impact so severely on deaf people. The effect of the cap in both its original and revised form falls almost entirely on deaf people who are, of course, already disadvantaged by their disability.
96. There is already evidence of prejudice caused to them by the cap, leading some to work "non-communication" days and putting strain on their relations with their employers. There are no exceptions, no discretion to waive the cap and no assessment is made of whether to require the employer to "plug the shortfall". The cap will tend to drive deaf people out of managerial jobs and into manual and unskilled jobs requiring lower levels of support, she argued.
97. For the Secretary of State, Ms Leventhal accepted the challenge of proving that the cap was justified even though it has a disparate adverse impact on the affected deaf people, including Mr Buxton himself. She submitted that when applying the three stage test of justification to a measure such as the cap, a legitimate social policy objective decided upon by a government body in the area of social and employment policy, must be held to qualify as a legitimate aim: Brachner v. Pensionsversicherungsanstalt Case C-123/10, CJEU, at [70].
98. Further, she submitted that "in choosing the measures capable of achieving the aims of their social and employment policy, the Member States have a broad margin of discretion" (ibid. at [73]).; and that the measure must be held to be proportionate will be met if the aim is "unrelated to any discrimination" and that the decision maker "could reasonably take the view that the means chosen were suitable for attaining that aim" (ibid. at [74]).
99. This approach is similar to that of the common law and of domestic law applying article 14 of the Convention; "great weight" should be placed on the judgment of those charged with administering a benefits system as an instrument of social policy, since government departments and public authorities, not the courts, are the "experts in administration" (per Lady Hale PSC at [36(6)] in R (C) v. Secretary of State for Work and Pensions [2017] 1 WLR 4127); cf., in the context of indirect discrimination, at [44]).
100. It was wrong, said Ms Leventhal, to describe the cap as merely a cost saving measure. The budget remained the same, having increased in real terms year on year since 1994, but awards would be spread among more people. The other aims were to influence the behaviour of employers, provide better value for money and ensure fairness in distributing the finite funds available. These aims go beyond just saving costs.
101. Next, she submitted that there is a rational connection between the level at which the cap is set and the achievement of these aims. The increase from 1.5 to 2 times average annual salary reflects representations made by Mr Buxton and others of like mind, while furthering the objective of spreading the fruits of the Scheme to greater numbers of disabled people. This strategy is intended to enhance equality of opportunity and foster good relations.
102. As for proportionality, the appropriate margin of appreciation must be accorded to the minister, said Ms Leventhal. The judgment whether to have a cap and how high to set it, was primarily for her. The revised cap was set at a level that reduced considerably the number of those affected. It was "benchmarked" by reference to prevailing NUBSLI rates. In cases where a second BSL interpreter might be needed, this could be a reasonable adjustment for the employer to provide.
103. The revised cap was accompanied by a host of "mitigation measures" announced in the statement to the House: guidance on best practice, continued monitoring of the impact of the cap, discretion in exceptional cases, managed personal budgets, acceptance of applications ahead of the starting date in a new job, investment in digital improvements to the application process, increased flexibility to support short periods of work experience in some cases and encouragement of technological solutions on a "risk free" trial basis.
104. I come briefly to my reasoning and conclusions on this second ground of challenge. I allow Mr Buxton to amend his claim to enable him to advance this ground. I find that it is arguable. All the elements for a successful indirect discrimination claim are present, subject only to the issue of justification, which it is the Secretary of State's burden to establish. The issue of justification is arguable. I grant permission for the indirect discrimination ground of challenge to proceed.
105. I do not, however, find that the second ground of challenge succeeds. The arguments of the Secretary of State are compelling in favour of justification. It is quite wrong to characterise the cap as purely a cost saving measure. There is no change to the budget, nor is there any ceiling to be placed on the overall pot of money, which will continue to be determined by the Treasury and set at a level commensurate with the estimated demand on the Scheme.
106. The changes made when the first cap was introduced in 2015 have not led to any major change in the level of expenditure on the Scheme each year. They were not intended to do so and have not done so in fact. The cap was introduced as an intended cost-neutral measure, changing the distribution of available funding but not its overall level. The evidence is clear that the increase in the level of the cap from April 2018 is intended to operate in the same way, while tilting the balance back again somewhat in favour of the high cost award recipients.
107. Second, I have no doubt that the aims of the Secretary of State in setting the cap and then revising it upwards, were and are legitimate aims. The aims are instruments of social and employment policy: to bring more disabled people into the workplace, improve compliance by employers with their duties to provide adjustments and distribute the Scheme's funds in a manner that is judged fair as between different categories of recipients. These aims are self-evidently legitimate.
108. Next, it is clear to me that the adjusted cap decided upon in March 2018 has a rational connection with the achievement of those aims. Put simply, the 2018 cap strikes a balance, differently from the balance struck in 2015 but a balance struck nonetheless, between potential new low cost award recipients, for whom money will be freed up, and existing high cost award recipients, who must perforce forego some of the available funds to provide the money to spend on achieving the government's aims.
109. I am in no doubt that the decision to set the 2018 cap at twice annual average earnings passes the test of proportionality. The court cannot judge how the balance of interests should be struck. That is the minister's job. The rationale is, however, there to be seen. The sacrifice demanded of high cost award holders, mostly deaf persons such as Mr Buxton, is pitched at a level that still leaves them enough money to pay for most or all of a BSL interpreter's services for 230 days each year at NUBSLI rates.
110. I cannot see any reason why the minister should not reasonably take the view that the means chosen was suitable for achieving the government's aims. The loss of income to Mr Buxton and other deaf people is less than under the more draconian 2015 cap and is, to some extent, tempered by the mitigation measures mentioned by the minister in her announcement and relied on by Ms Leventhal in her submissions.
111. The Secretary of State therefore succeeds in showing that the measure chosen was a proportionate means of achieving a legitimate aim, within section 19(2)(d) of the 2010 Act. The indirect discrimination ground of challenge, though arguable, is defeated by the success of that defence.

Conclusion

112. For those reasons, I refuse permission to amend the claim to advance the first ground of challenge (breach of the public sector equality duty); I grant permission to amend the claim to advance the second ground of challenge (indirect discrimination), and for the amended claim to proceed on the second ground; but the substantive claim fails and is dismissed.