A Booth v Delstar International Ltd

As I am considering both an appeal and a cross-appeal, I refer to the parties as the ‘claimant’ and ‘Delstar,’ to avoid confusion. This is a full hearing of both appeals against the judgment of an Employment Tribunal (“ET”), sitting in Leeds and chaired by Employment Judge Deeley and sent to the parties on 19 October 2020. The ET had allowed one claim of discrimination arising from disability, contrary to section 15 of the Equality Act 2010 (“EqA”) and indirect discrimination contrary to section 19 EqA in relation to Delstar’s delay in applying for income protection benefit. I refer to them for convenience as the “delay” claims. The ET had dismissed all of the other claims, including a claim under section 15 EqA that Delstar had attempted to dismiss the claimant, at a meeting on 5 February 2018. I refer to that claim as the “dismissal” claim.

The claimant appealed against the ET’s decision, which this Tribunal received on 7 December 2020. Permission to appeal was initially refused on the papers, but following a hearing under rule 3(10) EAT Rules 1993 (as amended), Gavin Mansfield KC, sitting as a Deputy Judge of the High Court, allowed one ground of the appeal to proceed, in orders dated 8 September 2021, which I set out later in these reasons. The ground relates to the ET’s decision on the dismissal claim.

Delstar lodged an answer to the appeal and cross-appealed on 24 November 2021 against the ET’s decision on the delay claims. HHJ Shanks granted permission on the papers for the cross-appeal to proceed, in a decision dated 12 January 2022.

I refer to a core bundle prepared by the claimant, “CB”.

The claimant has been disabled, following a pulmonary embolism and kidney disease, since he suffered a life-changing stroke in late March 2017. He remains employed by, but has not returned to work for, Delstar since his stroke.

At the core of this appeal is how Delstar treated him between November 2017, following the start of his illness and September 2018, when Delstar’s insurers, Unum, confirmed that the claimant was covered by an income protection policy. During this period, Delstar sought advice from its occupational health (‘OH’) advisors, and eventually Unum, about the likelihood of the claimant’s return to work, for the purposes of considering the claimant’s continuing employment and his entitlement to income protection, after his company sick pay ran out. Initially, the claimant’s prognosis was unclear. However, the OH advisors advised in a report written sometime in December 2017 that the claimant was unfit for work, and was not likely to be fit in the foreseeable future.

The relationship between the claimant and Delstar deteriorated in early 2018. This was said to be because of how one of Delstar’s HR managers, a Ms Davis, viewed the claimant’s entitlement to benefit from the Unum policy. She erroneously believed that the claimant could not benefit from the Unum policy, as she believed that only those employees who were likely to recover could benefit, whilst those employees who were assessed as permanently unfit for work could not. She held that belief without having checked the policy, or asking Unum for clarification. As a result, she began conversations with the claimant about his dismissal or ill-health retirement, with an ex gratia payment, and she delayed applying on his behalf to Unum for income protection, which was available after 26 weeks’ absence. The focus of the dismissal claim is a meeting which began on 30 January 2018 and recommenced on 5 February 2018, at which she discussed with the claimant the possibility of his dismissal. Delstar informed the claimant of Ms Davis’s error, in or around late April/ early May 2018. Delstar then successfully applied for income protection, and the claimant received back-dated benefits on 14 September 2018.

The claimant presented a claim form on 20 December 2019, a copy of which starts at page 45 CB. These included the delay claims and the dismissal claim, which are the focus of this appeal. Delstar resisted all of the claims in a response, a copy of which starts at page 71.

I set out below only the relevant parts of the ET’s judgment. The parts in bold or underlined were so emphasised in the original.

At page 1 CB, the ET decided:

The ET identified a list of issues, the relevant passages of which included:

The ET considered indirect discrimination at paragraph 4:

The ET went on to cite the law, including in relation to discrimination arising, at paragraph 25:

The ET’s relevant findings on the dismissal claim were as follows:

The ET’s findings and conclusions on the section 15 delay claim were as follows:

The ET went on to conclude that the unfavourable treatment in the delay claim was not proportionate and upheld the complaint at paragraph 125.

The ET’s conclusions in relation to the dismissal claim are at paragraphs 133 to 136:

The ET’s relevant findings on the Section 19 EqA claim (indirect discrimination) were as follows:

The claimant’s amended ground which has been permitted to proceed is relatively brief and so is set out in full, (see page 39 CB):

Given its length, I only summarise Delstar’s cross-appeal. The gist is as follows. In relation to the section 15 delay claim, Delstar accepts that the ET was entitled to conclude that the delay was unfavourable treatment, but says that the ET was wrong to attribute the cause of that delay to the ‘something arising,’ namely the appellant’s long term absence (see question b, after paragraph 119 of the ET’s Judgment). The ET had identified at paragraph 121 that the delay was due to Mr Davis’s belief, but failed to consider, in reaching its conclusion at paragraph 122, whether the reason for the delay was caused by the ‘something arising’. Instead, Delstar argued that the operative cause was Mrs Davis’s mistaken belief, rather than the background facts of the claimant’s disability and absence.

In the alternative, if the ET’s conclusion on the dismissal claim was inconsistent with the delay claims, it was the decision on the dismissal claim which was perverse, particularly given the brevity of the ET’s reasons for its decision on the dismissal claim.

Delstar also submitted that the ET had erred in upholding the indirect discrimination delay claim, at paragraph 132, based on its reasoning that there was a PCP which put disabled persons at a particular disadvantage when compared to “non-disabled” persons, as the comparison required a group disadvantage faced by those who shared the claimant’s disability, and an application of the PCP to people without the claimant’s disability.

The representatives agreed that the ET’s conclusions on the section 15 claims could not both be correct. While it might obviously be possible, in principle, for two separate claims to have alternative outcomes, there was no principled basis to distinguish between the causes of the delay and dismissal claims, when they were both in the context of the same belief. Having initially suggested that were I to find an error of law in either one of the decisions, this Tribunal ought to remake them, both representatives urged me to remit any remaking back to the ET, despite the passage of time since the original decision.

For the claimant, Mr Croxford argued that the ET had fallen into error by focussing solely on paragraph 25 of the Presidential decision of Sheikholeslami v University of Edinburgh [2018] IRLR 1090, instead of Simler P’s fuller guidance at paragraph 31 of Pnaiser. The unfavourable treatment (dismissal) had been identified (as per paragraph 31(a) of Pnaiser). Paragraph 31(b) required Tribunals to determine what caused the treatment, or the reason for it. The focus at this stage needed to be on the mind of Delstar’s manager and an examination of her conscious or unconscious thought processes. There might be more than one reason and the ‘something’ that caused the unfavourable treatment need not be the main or sole reason, provided it had at least a significant influence on the unfavourable treatment so as to amount to an effective reason for or cause or it. Motives were not relevant (para 31(c)).

The second stage of considering whether the ‘something’ was ‘arising in consequence of’ the claimant’s disability could describe a range of links, or a chain of them, so that where there was said to be more than one consequence of a disability, there needs to be a robust assessment of whether something can properly be said to have arisen in consequence of the person’s disability (see paragraph 31(d) of Pnaiser). The test is an objective one, and does not involve any consideration of thought processes (paragraph 31(f)). The two stages could be answered in any order (see paragraph 31(i)).

Mr Croxford argued that the ET had failed to consider the manager’s reasons for threatening to dismiss the claimant. Her motive was irrelevant, but her reasons were not. These were because of the ‘something’ that the claimant was on long-term sickness absence and was permanently incapacitated. She must have believed the Unum policy capable of applying to some employees, and the only possible group were those on long-term absence but expected to return, e.g. someone with a badly broken leg. The facts, as found by the ET, were that the manager had initially intended to apply for income protection but had changed her mind when she was advised that the claimant was permanently incapacitated. The manager had drawn the distinction between those employees who were permanently disabled and those who were absent but would return. That was why she had threatened to dismiss the claimant, when previously, when the prognosis had been unclear, she had not, and had intended to make the insurance application. The change was the prognosis of permanent incapacity. That was the ‘something arising’. Pnaiser had discussed at paragraph 31(e) an example of where a bonus was withheld by one manager because of a warning by another manager for absence, which arose from disability. There was no problem in concluding that the claim was made out. The claimant’s case was equally clear-cut. The manager’s subjective belief in the justification of her action was her motive, an impermissible consideration.

Ms Mellor returned to the case before the ET. The claimant was now arguing a case that he had not claimed. He had not claimed that the ‘something arising’ was permanent incapacity, but had relied on long-term sickness absence. The manager had been supportive of the claimant when he was on long-term sickness absence, and had not threatened him with dismissal, or considered the same, because of that absence, but because of her inexperience and unfamiliarity with the Unum policy terms. In contrast, the claimant’s long-term sickness absence was one of a number of links in the cause of the delay in applying for income protection insurance, and the ET’s reasons on the delay claim were either inadequate or perverse.

There are aspects of both representatives’ submissions with which I agree. I accept Mr Croxford's submission that part of the ET’s error flowed from its failure to consider the fuller set of guidance provided at para 31 of Pnaiser, rather than the briefer recitation of Sheikholeslami. Both Presidential decisions are, of course, correct, but Pnaiser explores in more detail the two aspects of causation, namely the ‘because of’ stage involving A’s explanation for the treatment (and conscious and unconscious reasons for it) and the ‘something arising in consequence’ stage, which requires consideration of whether, as a fact, the something was as a consequence of the disability. They may be answered in any order, but analysis of both is required. Pnaiser also discusses two other issues – the problem of multiple causes in the ‘because of’ analysis, and a chain of links in the ‘something arising’ analysis.

In its analysis of the dismissal claim, the ET referred at paragraph 135 to it having arisen from the manager’s mistaken belief about the terms of the Unum policy alone. I accept Mr Croxford’s challenge that there is no analysis of possible multiple causes. However, I also accept Ms Mellor’s submission that there is a second error, which might have been identified more clearly, had the Pnaiser guidance been adopted. This is that the ET was not consistent in how it defined the ‘something arising.’ A lack of consistency meant that both stages of the section 15 analysis were undermined.

In the List of Issues, the ‘something’ was defined at item 2(b)(i) “the fact that he was absent due to long term sickness.” This was for the delay and dismissal claims (items 2(b)(i)(a) and (b)). It did not refer to permanent incapacity. The ET reiterated this, at para b, after para 119, where it stated “The claimant relies on the following as “something arising”…the fact that he was absent due to long term sickness.” At paragraph 120, the ET found that the manager’s delay was due to her view that the Unum scheme did not apply “to employees who were on long term sickness absence who were unlikely to return to work….” The two are not necessarily inconsistent or mutually exclusive, but they are also not the same. The ET then returned to the original definition of ‘something’ at paragraph 122. In relation to the dismissal claim, at paragraph 134b, the ET referred to the claimant’s reliance on “the following as something arising in consequence of his disability: the fact that he was absent due to long-term sickness was the reason why the respondent decided to consider terminating the claimant's employment at the meeting on 5 February 2018.” In contrast to the delay claim, the ET did not refer to the manager’s belief that the Unum policy did not apply to those employees who were unlikely to return to work, in the dismissal claim. Instead, the ET referred to the belief that “the claimant was not eligible for income protection benefit under the Unum scheme,” and stopped there, without referring to the likelihood of return to work. The ET did not explain why the same belief caused one unfavourable treatment, but not another, and how the findings on permanent incapacity related to the ‘something’ relied on for the purposes of the delay claim.

In summary, the ET erred in focussing on a single cause (the manager’s belief at the ‘because of’ stage) when analysing the dismissal claim, while making findings in the delay claim which did not correspond to the ‘something’ relied on. These two errors, in combination, explain why the ET reached different decisions, which the representatives accepted could not both be correct.

I have considered whether it is appropriate to preserve either of the ET's conclusions. I conclude that it is not appropriate to do so. The issue of causation needs to be addressed with findings on the common, accepted ‘something arising’, namely the claimant’s absence due to long term sickness absence, not permanent incapacity, and the ET needs to consider whether there are multiple causes. This is not a case where the facts only lend themselves to one conclusion. It is appropriate that the ET should consider each of these two claims afresh.

The claimant's appeal and Delstar's cross-appeal in relation to the section 15 EqA delay and dismissal claims both succeed. The ET’s conclusions on both are not safe and cannot stand. I remit both issues to the ET to consider again, as the representatives have urged me to.

Ms Mellor argues first, that there was no evidence that those sharing the claimant’s disability would be put to a particular disadvantage. There was no evidence to support group disadvantage and the ET had erred in comparing people with disabilities with “non-disabled” people, which was a comparison appropriate under section 20 EqA, (the duty to make adjustments) not section 19. The ET had also implicitly acknowledged that the PCP could not have applied to people who did not share the claimant’s disability, at paragraph 128, when it said that Delstar “would have applied this PCP [not applying for income protection for those who were permanently ill] to non-disabled persons, albeit it is difficult to envisage a non-disabled person in such circumstances.” The ET had also accepted that any employee who met the Unum criteria was “highly likely” to be regarded as having a disability. The ET did not answer the question of how any PCP could have been applied to “non-disabled” people, ie, how the PCP applied generally, or could apply generally, beyond a disadvantaged group. It was no answer to say that this would deprive the claimant of a remedy, as many claims were misformulated as indirect discrimination claims.

Mr Croxford accepted that, as confirmed in Ryan v South West Ambulance Service NHS Trust [2021] ICR 555, for an indirect discrimination claim, the causative link needed to be not between the protected characteristic and the disadvantage, but the PCP and the group and individual disadvantage (see paragraph 31 of Ryan). However, the ET’s decision was consistent with that principle. The fallacy in Delstar’s argument could be seen by rephrasing the PCP, so that instead of not applying for income protection for employees who were permanently incapacitated, it was phrased as only applying for income protection for employees who are likely to be able to return to work. The second formulation had precisely the same substance and meaning and it was easy to see why it put the disabled group at a particular disadvantage and put the claimant at that disadvantage. On Delstar's own case, there was a group capable of benefitting from Unum, namely those employees who were on long-term sickness but not permanently incapacitated and it was for them that the manager had believed the Unum policy was in place. The general pool had been created by the manager’s mistaken belief, which put those who were disabled at a group disadvantage, and in turn the claimant to individual disadvantage. An example was of a person who had a particularly serious bone fracture, who was absent for long enough to benefit from income protection, (26 weeks), but less than 12 months. Ms Mellor had relied on section 6(3) EqA, namely those sharing the claimant’s characteristic needed to share the same disability, but all section 6(3) was intended to do was ensure that there was a comparison of like with like, within the disadvantaged group, in contrast with those who were not in the disadvantaged group. There was no need for statistical evidence of comparative disadvantage because of the nature of the factual case.

In contrast to section 15 claims, the PCP was stated as “not applying for income protection for those employees who were permanently ill, i.e. unlikely to return to work.”

I accept Ms Mellor’s submission that while the ET asked itself about two comparator groups, disabled and non-disabled people, that was not a correct comparison, as it reflected the section 20 EqA test, not the section 19 test. It is worth returning to the relevant statutory provisions. Section 19 EqA states:

Section 6(3) EqA states:

The meaning of section 19, when read with section 6(3), is clear. There is a requirement of a PCP of potentially general application, not just to those sharing a claimant’s disability. The group of people, of which the claimant is a member, must have the same disability as the claimant. There needs to a comparison between that group and those who do not share that disability. The comparator group may include those without any disabilities, and those with disabilities which are not the same as the claimant’s. The comparison is not between people with disabilities and those without, which is a different requirement under the section 20 duty to make adjustments (where a PCP puts a disabled person at a substantial disadvantage in relation to a relevant matter in comparison with persons who are not disabled). Both the claimant’s group and the comparator group may be hypothetical, because of the word, “would”, in subsections 2(a) and (b) of section 19. However, that does not avoid the need to analyse whether those in the claimant’s group would be put to a particular disadvantage (without falling into the trap of asking “why”, as it is often difficult to identify complex causes). While statistical evidence may not be available, the impact on one person, with the same disability, may not necessarily be the same as the impact on another. For example, it may be, depending on the evidence, that a proportion of those who share the claimant’s disability, a pulmonary embolism and kidney disease, would have still been viewed as likely to return to work, during the time period when the PCP was applied. While it is not a requirement that the PCP puts every member of the claimant’s group at a particular disadvantage, the ET had not asked that question, or gone on to consider whether those with the same disability would be viewed as more unlikely ever to return to work, when compared with those with different disabilities or no disabilities at all. That, ultimately, was the flaw in Mr Croxford’s reformulated PCP. It assumed that people with disabilities were more likely to be put to particular disadvantage than those without disabilities, without making a comparison between the two appropriate groups. As a consequence, the ET erred in its assessment of group disadvantage. It is no answer to point to a person who is not disabled, who had an absence lasting or likely to last more than 26 weeks (the deferral period under the Unum scheme) but less than twelve months, who would not be put to a particular disadvantage. While that person would be one member of the comparator group, it ignored other group members, including those with different disabilities, in comparison to those sharing the claimant’s disability. Contrary to Mr Croxford’s submissions, the ET’s analysis was not consistent with Ryan, because of the flaw in the ET’s analysis of group disadvantage. The ET will therefore need to revisit this analysis when assessing the section 19 EqA claim. I therefore also allow this part of Delstar’s cross-appeal against the ET’s decision on the section 19 claim.

I have considered the representatives’ representations that were I to find that the ET had erred in law, I should remit remaking back to the same ET if possible. There is no suggestion that it was biased or unprofessional in its conduct and despite some passage of time, the ET would not have to consider the whole case afresh. In the circumstances, having considered the authority of Sinclair Roche & Temperley & Ors v Heard & Anor [2004] IRLR 763, I regard it as appropriate to remit the matter to the same ET, if possible, or if not, to a differently constituted Tribunal.