Griffiths v The Secretary of State for Work And Pensions

The appellant became an employee of the respondent employer on 20 September 1976. She was at all material times employed as an Administrative Officer. From about October 2009 she started to experience the symptoms of conditions which were later diagnosed as post viral fatigue and fibromyalgia. As her employer subsequently conceded, this meant that she qualified as a disabled person within the meaning of the Disability Discrimination legislation. Following a 66 day absence from work, 62 of which were the result of an illness arising out of her disability, she was given a formal written improvement warning in May 2011. This was in accordance with the terms of the respondent’s Attendance Management Policy (“the Policy”) which envisaged the possibility of disciplinary sanctions after a specified period of absence from work.

The appellant raised a grievance in which she contended that due to her disabled status, her employer should in her case make two adjustments to the application of the Policy pursuant to section 20 of the Equality Act 2010. In broad terms this section specifies that where the arrangements adopted by an employer subject a disabled worker to a disadvantage, the employer should make reasonable adjustments designed to avoid the disadvantage. The adjustments she sought were these. First, she asserted that given that the lengthy absence which gave rise to the written improvement warning was the result of her disability, and moreover was when the disability was first diagnosed and a treatment plan put in place, the employer should not treat that absence as counting against her under the Policy with the consequence that the written warning should be withdrawn. Second, she wanted the Policy modified to allow her in future to have longer periods of illness absence before she faced the risk of sanctions than would be permitted for employees not subject to disability-related illnesses. She was unsuccessful in her grievance complaint and a further internal appeal was dismissed; consequently neither of the proposed adjustments was made.

She took a claim to the Employment Tribunal (“the ET”) alleging that the employer’s failure to make the adjustments she sought constituted a breach of the section 20 duty. (There is in fact no obligation on the employee to identify potentially reasonable adjustments but that was done here and they were the only potential adjustments considered by the Tribunal.) The majority dismissed her claims concluding that no duty to make either adjustment had arisen, and that in any event it was not reasonable for the employer to be expected to make either of them. The minority member would have upheld the claim on both grounds. An appeal to the Employment Appeal Tribunal (“the EAT”) was unsuccessful, the EAT (Mr Recorder Luba QC presiding) agreeing with the ET on both points. In addition the EAT held that the claim should fail for the further reason that the adjustments sought were not of a kind which fell within the terms of section 20. The appellant now challenges each of those conclusions.

The Policy provides that consideration will be given to formal action being taken against an employee where his or her absences reach an unsatisfactory level, known as “the Consideration Point”. This concept is explained in the introduction to the Policy as follows:

“The “Consideration Point” recognises that, as a human being, you are prone to illness and is a level of sickness absence within which you will not be subjected to formal action. It is set at 8 working days of sickness absence in any rolling 12 months … but may be increased as a reasonable adjustment if you are disabled.”

Paragraph 2 of the Policy provides that formal action will begin when the Consideration Point is reached.

The Policy itself, in paragraphs 2.3-2.4, spells out the purpose behind making adjustments to the Consideration Point for disabled employees and the circumstances where this may be appropriate:

2.3

Managers have a duty to make reasonable adjustments for disabled employees. Where appropriate, managers will allow a reasonable amount of additional sickness absence for a disabled employee when such absence is disability related. The purpose of increasing the Consideration Point in this way for disabled employees is:

•

To remove any disadvantage disabled employees may face by being expected to reach the same attendance standard as non-disabled employees

•

To ensure that disabled employees are clear about the attendance standard they are expected to meet and remove uncertainty about the possible consequences of taking time off as a result of their disability

•

To promote the continued employment of disabled employees

2.4

If the Consideration Point is increased it is known as the Disabled Employee’s Consideration Point. The Disabled Employee’s Consideration Point will be made up of the normal 8 days … for non-disability related absences and an additional number of days of absences related directly to the disability. Formal action will begin when:

Absences that are not related to the disability reach or exceed 8 working days; or,

The combination of disability-related and any non-disability related absences reach or exceed the Disabled Employee’s Consideration Point.

This means that whether an employee is disabled or not, formal action will begin at 8 days for absences unrelated to disability. But disabled employees have the flexibility to use these 8 days, (or some of them) as well as the additional number of days which has been agreed, for absences related to their disability if needed.

Paragraph 2.6 sets out what it terms “the formal stage for irregular absences”. There are four stages: an oral improvement warning; a written improvement warning; consideration of dismissal/demotion; and dismissal or demotion. The express reference to the consideration of a sanction as a separate stage 3 might suggest that whilst the first two stages are automatic once the relevant level of absence has been met, the final sanction is discretionary. In fact, however, paragraph 3 makes it clear that the imposition of these earlier warnings is also discretionary. It is provided that the managers should not give a warning if:

“(a)

One of the circumstances detailed in the list below applies …

The employee is disabled, the absence is directly related to the disability, and it is reasonable to increase the Consideration [Point]…

Taking into account the exceptional nature and/or circumstances of the absence andthe employee’s satisfactory attendance record, it would be perverse, unfair or disproportionate to give an improvement warning. …

(b)

They believe for reasons not detailed in the list that an Oral Improvement Warning would be inappropriate. This may include, for example, a reasonable expectation of improvement … ”

(The underlinings appear in the original document)

The adjustments sought by the appellant could, therefore, have been made consistently with the terms of the Policy. The written improvement warning given in response to the 62 day absence could have been revoked pursuant to para. 3(a) on the grounds of the exceptional nature and circumstances of that absence; and the additional permitted absence could have been achieved by establishing a new Disabled Employee’s Consideration Point as envisaged by para. 2.3 so as to confer the benefits described in para. 2.4.

The appellant first started to experience the symptoms of her disability in October 2009. Between October 2009 and 4 February 2010, she had 9 days of sickness absence. On 23 February 2010 her line manager, Mr. Dafydd Jones, gave her an oral improvement warning in accordance with the Policy. She does not complain about that warning.

Between 27 October 2010 and 2 May 2011 the appellant was absent from work for a total of 66 days. This was due principally (all but four days) to a continuous period of absence between 4 February 2011 and 2 May 2011 of 62 days. At an informal attendance review meeting on 7 March 2011 the appellant stated that her GP believed that she might be suffering from fibromyalgia and had referred her to a specialist. The Respondent subsequently received a report from the appellant’s consultant on 29 March 2011 which confirmed that the appellant had post-viral myalgia symptoms related to chronic fatigue but observed that her condition appeared to be gradually improving.

Following the appellant’s return to work on 3 May 2011, there was an attendance review meeting on 9 May 2011. She was given a written improvement warning. On 7 June 2011 her trade union representative submitted a grievance on her behalf. He asserted that the appellant was a disabled person within the meaning of the Equality Act 2010 and proposed that the two reasonable adjustments referred to above should be made. The grievance also requested that the appellant be referred to the Occupational Health Service to establish whether the condition from which she was suffering was a disability within the meaning of the Equality Act 2010. This last step was taken and the Occupational Health Adviser confirmed that the appellant would likely be considered a disabled person within the meaning of the legislation. Her disabled status has been conceded in this litigation.

On 7 November 2011 the appellant’s line manager, Mr Jones, heard and rejected her grievance. She pursued an appeal but this was dismissed on 29 November 2011.

The relevant law on disability discrimination is now, and was at the material time, found in the Equality Act 2010 which replaced provisions formerly contained in the Disability Discrimination Act 1995. The Equality Act gives effect to the United Kingdom’s EU obligations under Directive 2000/78/EC (“the Equality Directive”) which establishes a general framework for equal treatment in employment and occupation. The principle of equal treatment prohibits discrimination on certain prescribed grounds, such as race, sex and sexual orientation, and these include disability.

The structure of the Act is that it defines the meaning of discrimination and then identifies the circumstances when it is unlawful. There are four forms of disability discrimination: direct discrimination (section 13), indirect discrimination (section 19), discrimination arising from disability (formerly called disability-related discrimination) (section 15), and a failure to make reasonable adjustments (section 20). It is only the last of these which is specifically relied upon in this case. However, it is necessary briefly to touch upon the relationship between these various forms of discrimination in order to put into context the issues raised in this appeal.

The direct and indirect discrimination principles apply to all the prescribed characteristics, but the reasonable adjustments duty and the duty not to commit discrimination arising out of disability are unique to disability cases. They are a recognition of the fact that the difficulties faced by disabled workers are very different to those experienced by people subjected to other forms of discrimination. The point was succinctly expressed by Baroness Hale in Archibald v Fife Council [2004] UKHL 32; [2004] IRLR 651, para.47:

“But this legislation is different from the Sex Discrimination Act 1975 and the Race Relations Act 1976. In the latter two, men and women or black and white, as the case may be, are opposite sides of the same coin. Each is to be treated in the same way. Treating men more favourably than women discriminates against women. Treating women more favourably than men discriminates against men. Pregnancy apart, the differences between the genders are generally regarded as irrelevant. The 1995 Act, however, does not regard the differences between disabled people and others as irrelevant. It does not expect each to be treated in the same way. It expects reasonable adjustments to be made to cater for the special needs of disabled people. It necessarily entails an element of more favourable treatment.”

These two duties unique to discrimination therefore secure more favourable treatment and are closely interrelated. In GeneralDynamics Information Technology Ltd v Carranza [2015] ICR 169, HH Judge David Richardson succinctly explained the relationship between them (para. 32):

“The focus of these provisions is different. Section 15 is focussed upon making allowances for disability: unfavourable treatment because of something arising in consequence of disability is prohibited conduct unless the treatment is a proportionate means of achieving a legitimate aim.  Sections 20-21 are focussed upon affirmative action: if it is reasonable for the employer to have to do so, it will be required to take a step or steps to avoid substantial disadvantage.” 

Section 20, which is relied upon in this appeal, requires affirmative action in certain situations. It is headed “Duty to make adjustments”. So far as is material, it is as follows:

“(1)

Where this Act imposes a duty to make reasonable adjustments on a person, this section, sections 21 and 22 and the applicable Schedule apply; and for those purposes, a person on whom the duty is imposed is referred to as A.

(2)

The duty comprises the following three requirements.

(3)

The first requirement is a requirement, where a provision, criterion or practice of A’s puts a disabled person at a substantial disadvantage in relation to a relevant matter in comparison with persons who are not disabled, to take such steps as it is reasonable to have to take to avoid the disadvantage. …

(4)

The second requirement is a requirement, where a physical feature puts a disabled person at a substantial disadvantage in relation to a relevant matter in comparison with persons who are not disabled, to take such steps as it is reasonable to have to take to avoid the disadvantage.

(5)

The third requirement is a requirement, where a disabled person would, but for the provision of an auxiliary aid, be put at a substantial disadvantage in relation to a relevant matter in comparison with persons who are not disabled, to take such steps as it is reasonable to have to take to provide the auxiliary aid.”

Neither the second nor the third requirements are in issue here but they indicate the scope of the duty.

Schedule 8 to the Act deals with the reasonable adjustment duty in the context of work. Paragraph 2(b) provides that the reference to a disabled person is to an “interested disabled person” and by paragraph 4 (read with the appropriate table) that means an employee of the relevant employer. So the section 20 duty is owed by an employer to his employees. However, the duty does not arise if the employer does not know and could not reasonably be expected to know that the employee is disabled: sched.8 para. 20(1)(a).

The effect of section 21 is that a failure to comply with the section 20 duty to make reasonable adjustments amounts to an unlawful act of discrimination.

Focusing on section 20, it envisages that there will be a comparison of some kind between the disabled employee and those not disabled. The nature of the comparison was considered in the leading case of Archibald v Fife Council. It was not concerned directly with section 20 but with its statutory predecessor, section 6 of the Disability Discrimination Act 1995.The claimant suffered a disability following minor surgery which meant that she was not able to continue her work as a road sweeper. She applied for other jobs in the Council but was unsuccessful. Posts in the Council were awarded after holding competitive interviews, and she was not able to get to that stage. She complained that the Council ought to have made a reasonable adjustment by relieving her of the obligation to attend interviews and giving her a job for which she was suited. The issue before the House was whether that was in principle capable of constituting a reasonable adjustment. The Council argued that it was not and that it would require them to treat her more favourably than other employees. Their Lordships disagreed, holding that the reasonable adjustments duty could be engaged even where the employee was no longer capable of performing the duties she had been contracted to do. The duty could in principle extend to placing her in a different, even higher grade, post without a competitive interview if that was reasonable in all the circumstances. The case was remitted to the employment tribunal in order for the reasonableness question to be determined.

Lord Rodger observed that it was not altogether easy to discern the exact nature of the comparison required, and slightly different formulations were adopted by their Lordships when identifying the comparator in that case: see Lord Hope (para.12); Lord Rodger (para. 42); and Baroness Hale (para.64). But as Maurice Kay LJ pointed out in Smith v Churchill Stairlifts plc [2005] EWCA Civ 1220; [2006] ICR 524 para.39, there was no material difference in approach. Maurice Kay LJ suggested, and I would respectfully agree, that “the proper comparator is readily identifiable by reference to the disadvantage caused by the relevant arrangements” (para.39). In Archibald the comparator was an able bodied person who was not at risk of dismissal because, unlike Mrs Archibald, he or she was able to carry out the duties of the job. As Lord Rodger pointed out (paras 42-43), the substantial disadvantage was that Mrs Archibald was at risk of dismissal, and the purpose of the reasonable adjustment was to prevent the terms of Mrs Archibald’s contract from placing her at that substantial disadvantage. Allowing her to take a post, even possibly at a higher grade, without a competitive interview was capable of achieving that objective. In Archibald the disadvantage, namely risk of loss of employment, was very significant. But lesser prejudice will suffice. In the context of section 20 the word “substantial” is simply defined as “more than minor or trivial”: section 212(1).

Section 15(1) concerns discrimination arising out of disability and requires the employer in certain circumstances to make allowances for disabled employees:

“A person (A) discriminates against a disabled person (B) if -

(a)

A treats B unfavourably because of something arising in consequence of B’s disability, and

(b)

A cannot show that the treatment is a proportionate means of achieving a legitimate aim.”

This would mean, for example, that the dismissal of an employee for absences which are disability-related would amount to unfavourable treatment arising in consequence of disability and would therefore require justification.

Discrimination arising out of disability is in turn closely related to the concept of indirect disability discrimination which is defined in section 19. The concept of indirect discrimination is well established in discrimination law and it extends to those who are disadvantaged because they are disabled. Whereas direct discrimination under section 13 of the Equality Act occurs where someone is treated less favourably by reason of a protected characteristic (which includes disability), indirect discrimination is concerned with a provision, criterion or practice (“PCP”) (language echoed in section 20(3)), the effect of which is adversely to impact upon a particular group because of their shared characteristic. As with discrimination arising out of disability, but unlike direct disability discrimination, this adverse impact can be justified where the employer can show that the PCP is a proportionate means of achieving a legitimate aim.

I would draw attention to three matters with respect to these provisions. First, the definition of discrimination arising out of disability does not involve any comparison with a non-disabled person; it refers to unfavourable treatment, not less favourable treatment. The formulation of the duty prior to the Equality Act, in the Disability Discrimination Act 1995, did envisage such a comparison. In Lewisham London Borough Council v Malcolm [2008] UKHL 43; [2008] 1 AC 1399, a case I consider below, the House of Lords construed the relevant provision then in force so as effectively to make this form of discrimination a dead letter, in practice adding nothing to the concept of direct discrimination. The reformulation of the duty in section 15 of the Equality Act was designed to restore the law as it had been understood prior to Malcolm and thereby give the concept and the protection it affords real substance.

Second, it is perfectly possible for a single act of the employer, not amounting to direct discrimination, to constitute a breach of each of the other three forms. An employer who dismisses a disabled employee without making a reasonable adjustment which would have enabled the employee to remain in employment - say allowing him to work part-time - will necessarily have infringed the duty to make adjustments, but in addition the act of dismissal will surely constitute an act of discrimination arising out of disability. The dismissal will be for a reason related to disability and if a potentially reasonable adjustment which might have allowed the employee to remain in employment has not been made, the dismissal will not be justified. Finally, if the PCP, breach of which gives rise to the dismissal, also adversely impacts on a class of disabled people including the claimant, the conditions for establishing indirect discrimination will also be met.

Third, it is in practice hard to envisage circumstances where an employer who is held to have committed indirect disability discrimination will not also be committing discrimination arising out of disability, at least where the employer has, or ought to have, knowledge that the employee is disabled. Both require essentially the same proportionality analysis. Strictly, in the case of indirect discrimination it is the PCP which needs to be justified whereas in the case of discrimination arising out of disability it is the treatment, but in practice the treatment will flow from the application of the PCP. Accordingly, once the relevant disparate impact is established, both forms of discrimination are likely to stand or fall together. However, the converse is not true. If it is not possible to establish that the relevant PCP created a disparate impact, the case will not fall within the concept of indirect discrimination but it may nonetheless constitute discrimination arising out of disability.

A statutory Code of Practice on Employment has been published by the Equality and Human Rights Commission. Courts are obliged to take it into consideration whenever it is relevant: section 15(4). Chapter 6 is concerned with the duty to make reasonable adjustments. Paragraph 6.2 states:

“The duty to make reasonable adjustments is a cornerstone of the Act and requires employers to take positive steps to ensure that disabled people can access and progress in employment. This goes beyond simply avoiding treating disabled workers … unfavourably and means taking additional steps to which non-disabled workers … are not entitled.”

Paragraphs 6.23 to 6.29 of the Code give guidance as to what is meant by “reasonable steps” and paragraph 6.28 identifies some of the factors which might be taken into account when deciding whether a step is reasonable. They include the size of the employer; the practicability of the proposed step; the cost of making the adjustment; the extent of the employer’s resources; and whether the steps would be effective in preventing the substantive disadvantage. So far as efficacy is concerned, it may be that it is not clear whether the step proposed will be effective or not. It may still be reasonable to take the step notwithstanding that success is not guaranteed; the uncertainty is one of the factors to weigh up when assessing the question of reasonableness: see the observations of Lewison LJ in Paulley v First Group plc [2014] EWCA Civ 1573; [2015] 1 WLR 3384, paras 44-45.

Before the ET it was common ground that there was a relevant PCP as defined by section 20(3). The ET defined it as “a requirement to attend work at a certain level in order to avoid receiving warnings and a possible dismissal.” The appellant repeated the submissions she had made to the employer: that the 62 day absence should have been treated as an exceptional absence and disregarded for the purposes of the Policy with the consequence that the written warning would be revoked; and that the consideration point should be extended. It was suggested that there should be twelve days added to the eight already conferred upon all employees, thereby giving what the Policy refers to as a Disabled Employee’s Consideration Point of twenty days in all.

The appellant was not complaining about the Policy itself; it allowed for adjustments where appropriate. Her complaint was the application of the Policy in her particular circumstances.

The Secretary of State submitted that the duty to make adjustments did not even arise. Section 20(3) requires that the disabled person is placed at a substantial disadvantage when compared with an able bodied comparator. That was not so here; the Policy applied the same rules to everyone, disabled and non-disabled. The former were not placed at a substantial disadvantage. Further, the Secretary of State submitted that even if the duty was engaged, the proposed adjustments were not reasonable in the circumstances.

The majority of the ET accepted these submissions. They observed that:

“The policy applies to all. All face the same consequences if the absence level triggers a response under the policy…. The policy did not put the Claimant at a substantial disadvantage compared to a non-disabled person so far as sanctions were concerned.”

Earlier the ET had said that “the claimant is in a neutral position and even if not at an advantage she was certainly not at a disadvantage.”

The majority relied heavily upon the decision of the EAT in Royal Bank of Scotland v Ashton [2011] ICR 632 for the proposition that a policy applicable equally to all would not engage the reasonable adjustments duty. The majority noted that the substantial disadvantage relied upon was “worry and stress and the threat of losing the job”. But it concluded that the non-disabled would be similarly affected. The appellant was not therefore subject to a substantial disadvantage and the claim failed as a matter of law. The duty to make adjustments simply did not arise.

However, the majority did go on to consider the question of reasonableness and held that even if the duty was in principle engaged, neither of the proposed adjustments was reasonable. The reasoning was as follows (paras 47-49):

“47.

Further the reasonableness of the adjustments sought would inevitably rely upon the removal of almost a 3 month period of 66 days and then separately to allow a buffer to be continually in place which would be continually a factor in preventing the consideration point arising. Further, the majority do not think that the one-off exceptional circumstances category would apply given the condition of the appellant and the medical report as to the likelihood of continuing disability causing absence.

48.

We agree at Paragraph 28 of the respondent’s submissions that this would be in practice a perpetual extension of sickness absence not to assist the appellant to remain at work, though still employed.

49.

A further practical consideration arises out of the decision of O’Hanlon v Revenue & Customs Commissioners which in our view is correctly identified by the respondent. Namely that it be invidious for an employer to have to determine the extent to which sickness absence consideration points should have to be extended as part of a reasonable adjustment dependent upon any particular employees level of anxiety and/or stress, and/or fear of disciplinary action.”

The minority member dissented on both grounds for the following reasons (para. 52):

“The minority view is that the respondent’s sickness absence policy did place the appellant at a substantial disadvantage in comparison with non-disabled persons. Although the policy applies to all, disabled workers are more likely than non-disabled workers to have a high level of sickness absence, and the appellant was more likely than her non-disabled colleagues to reach the consideration point of 8 because of the nature of her disability, namely Post-viral fatigue and Fibromyalgia. In relation to the question of whether it was reasonable for the respondent to make the adjustments to the policy that the appellant requested, which were (a) to disregard the 62 days absence as a ‘one-off’ and (b) to increase the consideration point from 8 to 12, the following two arguments are persuasive to the minority. Firstly with regard to (a) the appellant’s representative informed the Tribunal that it was during this long absence that the appellant finally had a diagnosis, treatment was put in place and from there on in she could better manage her condition. Secondly with regard to (b) the respondent is a large organisation (para 6.28 of the EHRC Employment Code) and could reasonably accommodate an increased consideration point from 8 to 12. It is accepted that making these two reasonable adjustments would have removed the substantial disadvantage to the appellant of an increased likelihood of disciplinary action and dismissal and the consequent worry and stress. The appellant also gave evidence to the Tribunal that stress exacerbates the symptoms of her illness.”

(It is accepted that the minority was in error in saying that the consideration point would move from 8 to 12 days. The proposal was that it would be extended by 12 days giving a Disabled Employee’s Consideration Point of 20 days.)

The EAT rejected the appeal. The EAT agreed with the ET that no duty to make adjustments had arisen. It held that the conclusion that the appellant had not been placed at a substantial disadvantage was entirely consistent with the decision of the EAT in Royal Bank of Scotland v Ashton [2011] ICR 632. That case also concerned a sickness absence scheme which, as in this case, permitted the scheme to be modified in favour of disabled employees. The EAT in Ashton held that there had been no disadvantageous treatment in these circumstances and therefore no duty to make reasonable adjustments.

In view of its conclusion on the duty point, the EAT only touched very briefly upon the question of the reasonableness of the proposed adjustments. It said that the majority of the ET “had been entitled to find, as a matter of fact, that neither was reasonable.”

The EAT went further than the ET in accepting a submission from the Secretary of State that the adjustments sought, being concerned with past and future absences from work, did not fall within the concept of “steps” within the meaning of section 20(3). They were not, therefore, within the scope of the statute. For this reason also it was strictly unnecessary for the EAT to determine whether the proposed adjustments were reasonable or not.

There was an additional argument advanced by the appellant, which the EAT rejected, to the effect that the majority in the ET had failed properly to engage with the submissions of the appellant, having misunderstood the nature of the proposed adjustments she was seeking. Again the EAT dismissed this ground in brisk fashion merely stating that it was “unable to accept that the majority did actually fail to grasp the way that the case on the adjustments was being put”.

The grounds of appeal challenge each of the EAT’s conclusions. Accordingly, there are four issues to be addressed. First, was the majority of the ET right to conclude that there was no substantial disadvantage so as to engage the duty to make reasonable adjustments? If that was a sustainable conclusion, the appeal must fail whatever the merits of the other points. Second, was the EAT right to conclude that the proposed amendments were not “steps” within the meaning of section 20(3)? If they were not, the issue of reasonableness does not arise. Third, if there was a duty and the proposed adjustments did constitute potential steps which might be taken, did the majority misunderstand, and therefore fail to engage with, the appellant’s submissions on reasonable adjustments? Finally, if the majority did properly engage with the argument, was it entitled to find that it was not reasonable to expect the employer to make either of the proposed adjustments?

In order to engage the duty, there must be a PCP which substantially disadvantages the appellant when compared with a non-disabled person. In this case the PCP was, in the words of the ET, “a requirement to attend work at a certain level in order to avoid receiving warnings and a possible dismissal.”

Both the ET and the EAT considered that the Policy applied equally to all in circumstances which gave rise to no disadvantage. Indeed, to the extent that the Policy permitted a more lenient application of the principles to disabled employees by permitting them longer periods of absence before the imposition of sanctions is considered, the Policy was potentially more favourable to disabled employees.

Central to the analysis of both the ET and the EAT was the authority of Royal Bank of Scotland v Ashton. That was in certain respects similar on the facts. The appellant suffered from severe migraines and it was accepted that she was disabled within the meaning of the Disability Discrimination Act. She had frequent, though intermittent, absences from work. As in this case, the sickness policy provided for certain absence trigger points which, if exceeded, could lead to disciplinary action. The trigger points could be relaxed and they were for Ms Ashton; she was absent for very lengthy periods in the years 2006 to 2007 and 2007 to 2008 without any action being taken against her. In May 2008 she was given a warning as to her attendance for 12 months and her sick pay was withdrawn with respect to subsequent absences. That was the standard response in accordance with the policy to workers whose sickness levels had led to a warning. She submitted that it would have been a reasonable adjustment for the employer to exercise the discretion to defer the application of the disciplinary procedure, and also not to have withdrawn sick pay, albeit that this was in line with the policy.

Mr Justice Langstaff P, giving the judgment of the EAT, rightly observed that when considering the question of reasonable adjustment, it is critical to identify the relevant PCP concerned and the precise nature of the disadvantage which it creates by comparison with its effect on the non-disabled. The importance of this is that until the disadvantage is properly identified, it is not possible to determine what steps might eliminate it. These observations of the President were approved by Laws LJ giving the judgment of the Court of Appeal in Newham Sixth Form College v Sanders [2014] EWCA Civ 734 para.9.

In Ashton the EAT was not satisfied that the precise nature of the disadvantage had been properly identified by the ET, not least because the employee in that case had already been allowed very considerable leeway before the sanctions were imposed. But in any event the EAT appears to have accepted the employer’s submission that in the circumstances it was not open to the ET to find that there was any disadvantage. The following paragraphs support this reading of the decision:

“The statute expressly calls for a comparison with persons who are not disabled. In this part of its decision there is no obvious discussion of such persons, let alone any categorisation of those who would form the relevant comparator class. Mr Linden QC would have it that the class of persons who are not disabled here would mean all other employees in the employment of RBS, subject to the Sickness Absence Policy but not disabled, that is employees going about their business day to day and not regularly sick, as well as those who from time to time might be. By comparison with them he says, in line with the overview that we have just recorded, it could not be said that the appellant could possibly be disadvantaged, rather the opposite.

Mr Morton would argue that the comparison here should have been with persons who were not sick by reason of disability. Those who were sick by reason of disability would necessarily be more likely to subject to the trigger points and more likely, therefore, to be exposed to a disciplinary hearing and more likely, therefore, to be subject to a loss of sick pay in consequence. We have little hesitation in thinking that in particular bearing in mind that any comparison here should be a comparison of those who but for the disability are in like circumstances (see Malcolm) that Mr Linden QC's submission is correct. In any event, Mr Morton cannot point us to the particular group with which the Tribunal had in mind to make the comparison just as he cannot point us to any identification by the Tribunal of the disadvantage which the appellant suffered but which they in comparison did not.” (Italics added.)

Mr Leach, counsel for the employer, relies heavily on this analysis. There are in my view two assumptions behind the EAT’s reasoning, both of which I respectfully consider to be incorrect. The first is that the relevant PCP was the general policy itself. If that is indeed the correct formulation of the PCP, then the conclusion that the disabled are not disadvantaged by the policy itself is inevitable given the fact that special allowances can be made for them. It may be that this was the PCP relied upon in the Ashton case. But in my view formulating the PCP in that way fails to encapsulate why a sickness absence policy may in certain circumstances adversely affect disabled workers – or at least those whose disability leads to absences from work. Moreover, logically it means that there will be no discrimination even where an employer fails to modify the policy in any particular case. The mere existence of a discretion to modify the policy in the disabled worker’s favour would prevent discrimination arising even though the discretion is not in fact exercised and the failure to exercise it has placed the disabled person at a substantial disadvantage.

In my judgment, the appropriate formulation of the relevant PCP in a case of this kind was in essence how the ET framed it in this case: the employee must maintain a certain level of attendance at work in order not to be subject to the risk of disciplinary sanctions. That is the provision breach of which may end in warnings and ultimately dismissal. Once the relevant PCP is formulated in that way, in my judgment it is clear that the minority member was right to say that a disabled employee whose disability increases the likelihood of absence from work on ill health grounds, is disadvantaged in more than a minor or trivial way. Whilst it is no doubt true that both disabled and able bodied alike will, to a greater or lesser extent, suffer stress and anxiety if they are ill in circumstances which may lead to disciplinary sanctions, the risk of this occurring is obviously greater for that group of disabled workers whose disability results in more frequent, and perhaps longer, absences. They will find it more difficult to comply with the requirement relating to absenteeism and therefore will be disadvantaged by it.

The second error, in my judgment, is to assume that the ruling of the House of Lords in the Malcolm case, which was concerned with the nature of the appropriate comparison under the old concept of disability-related discrimination, is applicable to the obligation to make reasonable adjustments. In order to explain why that comparison is inapt in the case of the adjustment duty, I must briefly analyse the nature and scope of the principle enunciated in Malcolm.

Malcolm was not an employment case but concerned the termination of a tenancy. The tenant suffered from schizophrenia. He sub-let his flat, contrary to the tenancy agreement, and the Council sought to determine his tenancy. He claimed that the reason he acted as he did was because of his illness and that the act of termination constituted disability-related discrimination contrary to section 24 of the Disability Discrimination Act 1995.

Section 22(3)(c) of the 1995 Act provided that:

“It is unlawful for a person managing any premises to discriminate against a disabled person occupying those premises - ... by evicting the disabled person, or subjecting him to any other detriment”.

Section 24(1) provided that, for the purposes of section 22, a person discriminates against a disabled person:

“if (a) for a reason which relates to the disabled person's disability, he treats him less favourably than he treats or would treat others to whom that reason does not or would not apply; and (b) he cannot show that the treatment in question is justified.”

The appellant’s argument was that he was treated less favourably because of his disability, and that his eviction was not justified.

On the assumption that it could be shown that the reason for terminating the tenancy was linked to the disability, one of the issues before the House of Lords was how the relevant comparison envisaged by section 24 should be made. Who were the “others to whom that reason does not or would not apply”? This had been considered in detail by Mummery LJ giving judgment in the Court of Appeal in Clark v Novacold [1999] ICR 95. He illustrated the two competing constructions by taking the example of a blind man who wished to take his guide dog into a restaurant which had a “no dogs” rule. Should the comparison be with an able-bodied man who wished to take his dog into the restaurant? If so, there would be no less favourable treatment because all are treated the same. The able bodied man too would be refused entry for the same reason, namely that he wished to take his dog into the restaurant. Or should the comparison be with an able-bodied man who did not need to take a dog into the restaurant and would not therefore be excluded? In that case there would be unfavourable treatment. In the context of Malcolm the first approach would require the comparison with an able bodied man who had sub-let, and the second with someone who had not sublet.

The problem with the first analysis was that it effectively rendered disability-related discrimination a dead letter and equated it for practical purposes with direct disability discrimination, as Lord Brown recognised in terms (para.114). The problem with the second analysis was that it effectively did away with the comparison exercise altogether, as all their Lordships accepted. It requires a comparison with persons to whom the reason for the treatment does not apply; logically the appellant will always be treated less favourably than such persons.

The Court of Appeal in Novacold had preferred the latter approach on the grounds that it was what Parliament had intended, but in Malcolm their Lordships held, by a majority on this point (Lords Bingham, Scott, Brown and Neuberger; Baroness Hale dissenting) that the former was the proper comparison. So in the view of the majority, the comparison is a like for like exercise; the comparator must be similarly placed to the disabled claimant in all relevant respects save for the disability. This is precisely what is required in direct discrimination cases.

As we have seen, the new formulation of this duty in section 15 of the Equality Act removes the need for a comparison altogether (although a comparison may well in many cases cast light on thequestion whether the treatment is unfavourable). In effect section 15 has resurrected the Novacold approach, as Mummery LJ pointed out in Stockton on Tees Borough Council v Aylott [2010] ICR 1278 paras. 67-69.

Mr Leach contends that nonetheless the EAT in Ashton correctly held that Malcolm remains good law with respect to the relevant comparison exercise to be undertaken with respect to the section 20 adjustments duty. He submits that an earlier EAT decision in Fareham College Corporation v Walters [2009] IRLR 998, in which Cox J said that a Malcolm like for like comparison exercise did not apply to the reasonable adjustment duty, must be treated as impliedly overruled by Ashton.

If the approach in Malcolm were adopted, it would mean that the court should be indifferent to the reasons for absence just as it would be indifferent to the reason why the blind man wanted to take a dog into the restaurant. The fact that both the blind and sighted man who wished to take a dog into the restaurant were subject to the same disadvantage was enough. Similarly here; the employer treats disabled and able-bodied the same on the basis of the length of their absence and need not be concerned why the absence has occurred.

I have no hesitation in rejecting that submission. First, the language of section 20 is very different from the language in section 24 of the Disability Discrimination Act 1995. The nature of the comparison exercise in the former case is clear: one must simply ask whether the PCP puts the disabled person at a substantial disadvantage compared with a non-disabled person. The fact that they are treated equally and may both be subject to the same disadvantage when absent for the same period of time does not eliminate the disadvantage if the PCP bites harder on the disabled, or a category of them, than it does on the able bodied. Of course, if the particular form of disability means that the disabled employee is no more likely to be absent than a non-disabled colleague, there is no disadvantage arising out of the disability. But if the disability leads to disability-related absences which would not be the case with the able-bodied, then there is a substantial disadvantage suffered by that category of disabled employees. Thereafter the whole purpose of the section 20 duty is to require the employer to take such steps as may be reasonable, treating the disabled differently than the non-disabled would be treated, in order to remove the disadvantage. The fact that the able-bodied are also to some extent disadvantaged by the rule is irrelevant.

Second, the suggested approach is inconsistent with the Archibald analysis. The fact that all employees were at risk of dismissal if they were unable to perform the contractual duties of the job did not eliminate the need to make such adjustments as were reasonable when the reason for the inability to work was connected with disability. Mr. Leach boldly submitted that Archibald was implicitly overruled by Malcolm, but that is an impossible submission to sustain.The House of Lords in Malcolm would have had to address Archibald head on and explain why it was departing from a decision so recently given. Archibald was in fact cited to the House in Malcolm and although there was reference to the reasonable adjustments duty both by Baroness Hale (para. 75) and Lord Brown (para.107), only the latter mentioned Archibald, and then only in passing as an example of a case involving the reasonable adjustments duty. There is no suggestion anywhere in the judgments that Archibald was affected by the court’s ruling.

Third, there are other authorities which show that the section 20 duty is not satisfied by treating all alike. In O’Hanlon v Revenue and Customs Commissioners [2007] ICR 1359 the Court of Appeal considered the effect of sick pay rules which allowed full pay for six months in any twelve month period, then half pay for a further six months in any four year period, and thereafter the pension rate of pay. The claimant was frequently absent because of her disability and claimed that the rules subjected her to a substantial disadvantage because the sick pay leave attributable to her disability absorbed the sick pay which would otherwise have been available for non-disability related absences. She said that she should receive full pay for all disability-related illnesses or alternatively that periods of disability-related absence should be discounted when assessing periods of absence for the purposes of the sick pay rules. The employer argued that there was no disadvantage since all were treated the same. Lord Justice Hooper giving a judgment in the Court of Appeal with which Sedley and Ward LJJ agreed, rejected that argument. He approved the following observations of the EAT in that case (Mr Justice Elias P):

“… there is no basis for saying that the tribunal erred in law in reaching this decision. It is plainly no answer to a claim of this kind to say that the same rules apply to all. The purpose of the provision is that disabled employees may be disadvantaged by common rules. Unlike other forms of discrimination, the employer may be obliged to take positive steps which involve treating the disabled employee more favourably than others are treated to remove or alleviate the consequences of the disability.”

More recently, in HK Danmark, acting on behalf of Ring v Dansk Almennyttigt Boligselskab [2013] IRLR 571; [2013] ICR 851, the Court of Justice of the European Union had to consider a provision of Danish law which had the effect that an employee who had been paid full pay when absent from work for more than 120 days in a twelve month period would be entitled to a shorter period of notice than someone who had been absent for fewer days. The issue was whether this constituted discrimination against the claimant disabled worker who triggered the rule in circumstances where his absence was the consequence of his disability. The court noted that the equality of treatment meant that there was no direct discrimination on grounds of disability, but nonetheless the rule created a prima facie case of indirect discrimination within the meaning of Article 2(2)(b) of the Directive (para.76):

“A worker with a disability is more exposed to the risk of application of the shortened notice period … than a worker without a disability. As the Advocate General observes in point 67 of her opinion, compared with such a worker, a worker with a disability has the additional risk of an illness connected with his disability. He thus runs a greater risk of accumulating days of absence on grounds of illness and consequently of reaching the 120 day limit …. It is thus apparent that the 120 day rule … is liable to place disabled workers at a disadvantage and so to bring about a difference of treatment indirectly based in disability ….”

In that case the question of disadvantage was being considered in the context of indirect discrimination. The Court held that there was a disadvantage and the question was whether the rule could be justified, a matter for the national court. Here we are concerned with the concept of disadvantage in section 20, but there is no basis for giving it a different meaning. In Ring a common rule relating to sickness absence was held to disadvantage disabled workers who suffered from disability-related sickness absences. In my judgment it is clear that the common rule applied in this case equally disadvantages disabled workers whose absence is disability-related. Nothing in Malcolm affects that conclusion. The only question therefore is whether there are reasonable steps which would remove that disadvantage.

For these reasons I accept the appellant’s submissions that both the majority in the ET and the EAT were wrong to hold that the section 20 duty was not engaged simply because the Policy applied equally to everyone. The duty arises once there is evidence that the arrangements placed the disabled person at a substantial disadvantage because of her disability. That in my judgment was unarguably the position here.

Counsel for the respondent submitted, as he successfully had before the EAT, that neither of the proposed adjustments constituted “steps” within the meaning of section 20. The argument turns on the purpose of the section 20 duty. Counsel said that it was to enable disabled employees to return to work or carry on working. The Code expresses the purpose as being to promote “access to and progress in” employment. That was not, counsel submitted, what the proposed adjustments would achieve. On the contrary, they would if anything put a premium on the appellant absenting herself from work since there would be less incentive to return.

In my judgment, there is no reason artificially to narrow the concept of what constitutes a “step” within the meaning of section 20(3). Any modification of, or qualification to, the PCP in question which would or might remove the substantial disadvantage caused by the PCP is in principle capable of amounting to a relevant step. The only question is whether it is reasonable for it to be taken.

In my view the proposed steps would, if taken, be capable in principle of ameliorating the disadvantage resulting from the operation of the Policy. Indeed, the first adjustment involving the discounting of the first 62 day disability-related illness absence would plainly have that effect because it would involve the withdrawal of the written improvement warning. That would reduce the risk of dismissal for further absences; or at least would delay it. Similarly the second might make it less likely that the appellant would be disciplined in future, although whether it would in practice have that effect would depend upon the length and frequency of future disability-related absences.

This conclusion is supported by the fact that in providing an illustration of what may be involved in the duty to make a reasonable adjustment, the Code takes the example of a person with auto-immune disease who has taken several short periods of absence relating to his disability. The Code suggests that it may be a reasonable step to discount these periods when taking absences into account in the context of a redundancy selection exercise. In my judgment, there is no reason in principle why such absences could not be discounted in the context of determining whether or not to dismiss or take other remedial action for absenteeism. It is also relevant in my opinion that the Policy itself envisaged that this might be an appropriate step for the employer to take with respect to disabled employees.

However, I would accept that whilst a disabled employee may suffer disadvantages not directly related to the ability to integrate him or her into employment, the steps required to avoid or alleviate such disadvantages are not likely to be steps which a reasonable employer can be expected to take. The O’Hanlon case, referred to above, provides an example. The sick pay rules in that case were found to constitute a substantial disadvantage for reasons already discussed, but the Court of Appeal agreed with the EAT in holding that increasing the period during which the disabled claimant could claim full pay whilst sick would not be a reasonable step to expect the employer to take. This was particularly so in that case because the reason why it was said that the employer should do this was because the claimant was personally experiencing financial hardship as a result of the reduction in her sick pay which she said caused her stress and exacerbated her illness. Hooper LJ held, adopting the reasoning in the EAT, that this was not a reasonable adjustment, not least because it would mean that the employer would have the invidious task of having to assess the financial means of his disabled employees and the stress suffered as a result of any hardship. More generally, Hooper LJ also approved an observation by the EAT that:

“The Act is designed to recognise the dignity of the disabled and to require modifications which enable them to play a full part in the world of work, important and laudable aims. It is not to treat them as objects of charity which, as the tribunal pointed out, may in fact sometimes and for some people tend to act as a positive disincentive to return to work.”

The appellant submits, as she did before the EAT, that in the discussion of reasonableness (see para. 35 above) the majority in the ET simply failed to discuss the two adjustments which the she was seeking. It did not engage with the way in which the case had been advanced. Mr Ford, counsel for the appellant, submits that the ET appear to have understood the appellant’s case to be that any disability-related absence should be ignored when applying the Policy, however long it may be, whereas the proposed adjustment was more limited. It did no more than seek to have the Policy adjusted in precisely the way the Policy itself envisaged, namely by allowing a longer period of absence before the consideration point was reached - in this case an extension of twelve days. Furthermore, when considering whether to ignore the first lengthy absence altogether, the tribunal had failed to have regard to the reason why that was said to be exceptional, namely that it was only during that absence that the nature of the disability had been diagnosed and appropriate treatment put in place so that the appellant could thereafter better manage her condition.

I agree that the majority does not in terms refer to the argument as advanced by counsel on the appellant’s behalf. Nonetheless, I do not accept that the ET failed to appreciate the way in which the case had been advanced. As to the first proposed adjustment, I read the ET as saying that the medical evidence did not support the inference that the original 66 day absence was simply a “one-off” absence, given the likelihood of continuing and potentially lengthy disability absences. We have seen the evidence. The most recent report from an occupational health advisor confirms that the problems are likely to recur and future absences are to be expected, although their frequency and pattern is difficult to predict. Nor was it suggested that this absence was unusually lengthy because of the need to diagnose the illness.This is why the majority said that it could not properly be considered as an “exceptional” absence within the meaning of the Policy, notwithstanding that the illness was only diagnosed then. It was not objectively reasonable to expect the Secretary of State simply to ignore such a lengthy absence. In my judgment, the majority would have been well aware that this was an important element in the argument being advanced, not least because it was the reason why the minority member dissented. It is inconceivable that the members would not have explored this issue in discussion.

As to the modification of the consideration point, it was clear that the appellant was not in fact asking for all disability-related absences to be ignored. The appellant’s representative made that clear in written submissions to the court, as the ET noted at paragraph 25. And again, the minority member decided the case in the appellant’s favour on that basis. I cannot believe that the majority was ignorant of, or misunderstood, the nature of the submission. I accept that the observation in paragraph 47 to the effect that the appellant was seeking to “allow a buffer to be continually in place which would be continually a factor in preventing the consideration point arising” might suggest otherwise. But in my view the ET was there accepting an argument from the employer to the effect that if the worry and stress were to be eliminated entirely, this could only be achieved if, in effect, all disability-related illness was ignored. That objective would not be met by a short increase in the period of absence allowed before the consideration point was reached. The majority clearly considered that this would be unreasonable and would not encourage a return to work.

I would therefore reject the submission that the appellant’s case was not properly understood or considered by the majority.

The question whether the proposed steps were reasonable is a matter for the ET and has to be determined objectively: see Churchill Stairlifts paras 44-45 per Maurice Kay LJ.The question for this court is whether that was a sustainable decision on the evidence before it.

In my judgment, it was open to the ET to find that these adjustments were not reasonable. I have had some concerns about whether the reasoning of the majority of the ET, at least on the second proposed adjustment, was sufficiently clear to justify its conclusions. But inadequacy of reasons, as opposed to a failure properly to consider the appellant’s case, was not a specific ground of appeal, and in any event I think that there is just sufficient explanation of the reasoning process to pass muster.

The ET does give relatively clear reasons why it did not consider it reasonable to expect the employer to ignore the original 62 day disability-related absence and revoke the written warning. The majority notes that this is not a one-off condition and that further periods of potentially lengthy absence would, on the medical evidence, be likely to arise. In that context, the majority of the ET did not consider that it was reasonable to expect the employer to write off an extended period of absence. As I have said, although the majority did not in terms address the argument that the absence was exceptional because it was the period when the illness was diagnosed and the treatment plan adopted, it would have had this in mind. It was in my judgment entitled to take the view that this was not a material reason for simply ignoring this lengthy absence. It was after all some eight times longer than the permitted annual absence before the consideration point is reached. In my judgment, the conclusion was plainly open to it.

In this context I would observe that it is unfortunate that absence policies often use the language of warnings and sanctions which makes them sound disciplinary in nature. This suggests that the employee has in some sense been culpable. That is manifestly not the situation here, and will generally not be the case, at least where the absence is genuine, as no doubt it usually will be. But an employer is entitled to say, after a pattern of illness absence, that he should not be expected to have to accommodate the employee’s absences any longer. There is nothing unreasonable, it seems to me, in the employer being entitled to have regard to the whole of the employee’s absence record when making that decision. As I mention below, the fact that some of the absence is disability-related is still highly relevant to the question whether disciplinary action is appropriate.

As to the second proposed adjustment, the reasoning of the majority is in my opinion more opaque. But I think implicit in its analysis is the belief that there is no obvious period by which the consideration point should be extended. If the worry and stress of being at risk of dismissal is to be eliminated altogether, then all disability-related illness must be excluded. But if that step is not taken - and no-one was suggesting that it should be - then in a case like this when lengthy further periods of absence are anticipated, the period by which the consideration point should be extended becomes arbitrary. As the majority point out in paragraph 49 when drawing an analogy with the O’Hanlon case, in so far as the alleged disadvantage is with the stress and anxiety caused to a particular disabled employee, it would be invidious to assess the appropriate extension period by such subjective criteria.

Also, where the future absences are likely to be long, a relatively short extension of the consideration point is of limited, if any, value. It will not in practice remove the disadvantage if the absences remain over 20 days. No doubt there will be cases where it will be clear that a disabled employee is likely to be subject to limited and only occasional absences. In such a situation, it may be possible to extend the consideration point, as the Policy envisages, in a principled and rational way and it may be unreasonable not to do so. But in my view the majority has taken the view that this is not appropriate in a case of this nature. In my judgment, the majority was entitled to reach that conclusion.

However, it is important to add this. As I have already discussed, the positive duty to make reasonable adjustments is only a part of the protection afforded to disabled employees. The fact that the employer may be under no duty to make positive adjustments for a disabled employee in any particular context does not mean that he can thereafter dismiss an employee, or indeed impose any other sanction, in the same way as he could with respect to a non-disabled employee. The employer is under the related duty under section 15 to make allowances for a disabled employee. It would be open to a tribunal to find that the dismissal for disability-related absences constituted discrimination arising out of disability contrary to section 15. This would be so if, for example, the absences were the result of the disability and it was not proportionate in all the circumstances to effect the dismissal.

This is particularly relevant to the first proposed adjustment. In substance the complaint is that it was disproportionate to impose the disciplinary sanction given that the absence giving rise to it was disability-related. It is that treatment which lies at the heart of the complaint, not the failure to make an adjustment. The section 20 duty is normally relevant when looking into the future; it is designed to help prevent treatment which might give rise to a section 15 claim from arising. But that is not the purpose of the section 20 complaint here. It is really a staging post in challenging in order to invalidate the written warning - treatment which has already arisen. In my view there is a certain artificiality in arguing the case in that way. I respectfully agree with some observations of HH Judge Richardson in General Dynamics Information Technology Ltd v Carranza [2015] ICR 169 para. 34 when he said that dismissal - and I would add any other disciplinary sanction - for poor attendance can be quite difficult to analyse in terms of the reasonable adjustments duty, and that:-

“Parties and employment tribunals should consider carefully whether the duty to make reasonable adjustments is really in play or whether the case is best considered and analysed under the new, robust, section 15.”

The same artificiality does not, however, relate to the second proposed adjustment. That is designed to look into the future and to limit the risk of future disciplinary treatment being meted out for absence from work which would be disproportionate. But even where there are no relevant reasonable adjustments of this nature to be made, the question would still arise, at the time of dismissal, whether the dismissal is a proportionate response to the pattern of absences having regard to all the circumstances, including the important fact that they may be wholly or in part disability related.

I would therefore dismiss this appeal on the single ground that the ET was entitled to hold that the proposed adjustments were not steps which the employer could reasonably be expected to take.

I agree.

I also agree.