Court of Appeal: Underhill VP, Baker and Davies LJJ, [2021] EWCA Civ 1703, 24 November 2021

This was an appeal from the refusal of a challenge to the lawfulness of the Self-Employment Income Support Scheme (“SEISS”) introduced by the government during the first Covid-19 lockdown. The claimants argued that the scheme breaches Article 14 ECHR read with A1P1 by discriminating against self-employed women who took a period of leave relating to maternity or pregnancy in any of the three relevant tax years on which SEISS payments were calculated, this because the level of support granted to them under the scheme was not representative of their usual profits. Whipple J had dismissed the claim having considered the extraordinary pressures under which the scheme was introduced (including the imperative to distribute funds speedily) and the fact that the scheme adopted operated on the basis of data already held by the state. She was not persuaded that the claimants had demonstrated indirect discrimination or Thlimmenos discrimination but proceeded to consider justification, upon which she found against the claimants having adopted the “manifestly without reasonable foundation” approach (the correctness of which had been common ground between the parties).

The claimants appealed on the basis that Whipple J had erred in her approach to indirect discrimination, to Thlimmenos-type discrimination, and to justification. The Court of Appeal (Underhill and Baker LJJ, with whom Davies LJ agreed) agreed that the Judge had misdirected herself as to indirect discrimination by failing properly to take into account the disparate impact of the scheme on women who had taken maternity leave. It found it unnecessary to consider the challenge to the Judge’s application of Thlimmenos and (having considered the decision of the Supreme Court in R (SC) v Secretary of State for Work and Pensions [2021] UKSC 26, [2021] 3 WLR 428 (see previous post) dismissed the appeal on the basis that Whipple J had been entitled to find that any discrimination was justified (further, that it was in fact so justified). The case provides further illustration (see also R (Salvato) v Secretary of State for Work and Pensions [2021] EWCA Civ 1482 and related post) that  the movement away from the “manifestly without reasonable foundation” test in cases where suspect grounds are in play is by no means a panacea for claimants.

The Court referred at §53 to the judgment of Lord Reed (with which the other members of the Supreme Court agreed) in SC §§49-53, as establishing the “most recent authoritative statement of the nature of indirect discrimination in the context of Article 14. It observed that the approach to indirect discrimination under Article 14 was “more generally expressed than the definition of indirect discrimination adopted in the EU legislation and jurisprudence” (§54), which required a “‘provision, criterion or practice’ … which puts persons with a protected characteristic at ‘a particular disadvantage’ compared with others”. Notwithstanding this, it stated at §56 that it was “legitimate in an Article 14 case at least to have regard to the EU and domestic jurisprudence” which was underpinned by the same principles (“the concept of a ‘measure’ does the same work as a ‘PCP’ in EU law, and the requirement that it has ‘disproportionately prejudicial effects’ (or ‘affects a disproportionate number of members’) on the relevant group essentially corresponds to the requirement that it puts members of that group at ‘a particular disadvantage’.” The Court cited the judgment of Lady Hale in Essop v Home Office [2017] UKSC 27, [2017] 1 WLR 1343  §§24-26, in which she stated that (1) “there was no “express requirement for an explanation of the reasons why a particular PCP puts one group at a disadvantage when compared with others… It is enough that it does”; (2) indirect discrimination did not require a causal link to be established between the less favourable treatment and the protected characteristic, rather “between the PCP and the particular disadvantage suffered by the group and the individual”; and (3) “the reasons why one group may find it harder to comply with the PCP than others are many and various … They could be genetic, …social, …traditional employment practices, … another PCP, working in combination with the one at issue….”

The Court ruled that Whipple J had erred in concluding (in reliance on the decision of the House of Lords in Barry v Midland Bank plc [1999] UKHL 38, [1999] 1 WLR 1465, that of the Court of Appeal in Trustees of Uppingham School Retirement Benefits Scheme v Shillcock  [2002] EWHC 641 (Ch), [2002] IRLR 702, and that of the Divisional Court in R (Adiatu) v HM Treasury [2020] EWHC 1554 (Admin), [2020] PTSR 2198 (see earlier blog) that there was no indirect discrimination because (§64) “The measure imposes no hidden barriers to eligibility. So far as quantum of payment is concerned there is no hidden barrier either: quantum is based on past (average) trading profits, which are a matter of past fact.” At §71 the Court of Appeal observed, as regards this conclusion, that this “picks up a similar statement in Adiatu … which itself takes the phrase from para. 25 of Lady Hale’s judgment in Essop …  but we are not sure what it means in this context”, concluding that “Even if it is in some sense true, it is not an answer to the claim that the adoption of the ATP measure has a disproportionate impact on recent mothers”.

Whipple J’s conclusion on indirect discrimination had also relied on her finding at §64 that “The same rule applies to all and it is no harder for a woman who has been on maternity leave to qualify or calculate their payment, than someone who has not. The fact that some claimants will receive lower payments than others reflects the fact of lower earnings in past years; I agree with the Defendant that the reasons for lower earnings in past years, in the context of this Scheme with its stated purpose, are not relevant.” At §72 the Court of Appeal ruled that the fact that “[t]he same rule applies to all” “cannot of course be an answer to the appellants’ case: the essence of prima facie indirect discrimination is that the same rule applies to everyone but that it has a disproportionate impact on the group in question. We are sure the Judge was well aware of that, and we assume she was merely emphasising that this was not a case of directly different treatment; but that in itself does not advance the argument”. And at §73, that ‘“it is no harder for a woman who has been on maternity leave to qualify or calculate their payment, than someone who has not’ is true, but it does not address the essence of the appellant’s case. The complaint is not that recent mothers are disqualified from receiving a SEISS payment, or that calculation of their entitlement is more difficult, but that the adoption of ATP as the measure of their entitlement puts them at a particular disadvantage”. Finally, as to the statement that “the reasons for lower earnings in past years, in the context of this Scheme with its stated purpose, are not relevant”, the Court at §74 stated that “the appellants’ point is not as such that recent mothers will receive lower SEISS payments than other self-employed people but rather that, compared with such people, they are at disproportionate risk of receiving payments which are unrepresentative of their hypothetical no-Covid earnings”. And while this “need not undermine the judge’s essential point, the essence of which is that the prejudice complained of is immaterial because it ‘reflects the fact of lower earnings in past years’” the conclusion that the reason for lower earnings in previous years “was immaterial” was inconsistent with the approach of Lady Hale in Essop (see §76).

Whipple J having reached her conclusions on the basis of the decisions in Barry, Shillcock and Adiatu, the Court proceeded to “consider those cases with some care”. It distinguished Barry, the only decision which was binding on it, having at §84 summarised the ratio (not without difficulty) as being to the effect that “It is essential to a claim of indirect discrimination that the group to which the claimant belongs should be differently treated from persons not in that group. In deciding whether there has been such a difference of treatment it is necessary to identify the true substance of the measure which gives rise to the claim, and that may involve a consideration of its purpose. The purpose of the scheme in Barry was to mitigate the impact on employees of the loss of the income that they would have received but for their dismissal. That being so, it cannot be regarded as a difference in treatment between part-time and full-time employees to leave out of account, in calculating the benefits payable, a factor (i.e. levels of earnings over the employee’s career as a whole) which was of its nature irrelevant to the loss which it was the purpose of the scheme to mitigate: what the part-time employees had lost was their part-time earnings… In terms of the overall structure of indirect discrimination …  the effect of the reasoning is that in such a case the claimant’s group is not prejudiced or “put at a particular disadvantage”: they have, in the material respects, been treated in the same way as everyone else”.

Having summarised the ratio of Barry thus, the Court of Appeal concluded at §87 that “it provides no analogy to the present case. The purpose of SEISS is to compensate self-employed persons for their loss of profits in the current year as a result of the pandemic. The ATP measure works by using past profits to represent, in however rough-and-ready a manner, their likely hypothetical no-Covid profits. If its use in the case of new mothers produces results which are disproportionately unrepresentative of those profits, as compared with others, that necessarily puts them at a particular disadvantage. By contrast, Mrs Barry’s previous (whole-career) earnings were irrelevant to the earnings that she would have received but for her dismissal”.

The Court found it unnecessary to deal with Shillcock other than briefly but considered Adiatu in detail. The Divisional Court’s there rejected the claim that the rate at which SSP was set indirectly discriminated against women and members of ethnic minorities who were “disproportionately represented in the lowest earning groups … disproportionately likely to be unable to have the resources to manage with such a low income, and … accordingly disadvantaged by the rate of SSP (either losing income or going to work when they ought not to do so)”. At §141 the Divisional Court had ruled that “[t]he rate of SSP is not a PCP which places certain categories of employees at a particular disadvantage”, that it was “not a barrier or gateway … [but] is a sum that is paid, in exactly the same way, to everyone who receives SSP, regardless of their protected characteristics. It does not place women or BAME employees at a particular disadvantage: everyone is treated the same.” At §149 the Court had gone on to rule that “Essop is not authority for the proposition that something places those with protected characteristics at a particular disadvantage because their circumstances, unconnected with the PCP, are less favourable than those of others… the Claimants do not rely upon any disadvantage that is caused by the rate of SSP itself. Rather, they rely upon an alleged disadvantage, the absence of other financial resources, which is not caused or related to the rate of SSP in any way. This does not turn the rate of SSP into a PCP which places women or BAME employees at a particular disadvantage.” The Court of Appeal at §91 agreed with the Divisional Court:

“The essential point being made is the same as that in Barry, although Mr Adiatu’s claim was in truth a good deal more ambitious than Mrs Barry’s. SSP is not a benefit the purpose of which is to mitigate the effects of low income, still less the effect of the pandemic on those with low income; and the fact that a claimant has low earnings is simply immaterial. That being so, the fact that a higher rate of SSP would be beneficial to those on low incomes cannot found an argument that they are materially disadvantaged by the actual rate…”.

At §92, however, the Court disagreed with Whipple J’s belief that Adiatu “supported her conclusion that new mothers’ earnings in the measurement period could not be relied on to found a case of particular disadvantage … we do not believe that it does so… The purpose of SEISS is to compensate self-employed people for the loss of the earnings that they would have received in the current year but for the pandemic and to use past earnings as the measure of those lost hypothetical earnings. In those circumstances, the past earnings in question are not immaterial: on the contrary, they are crucial”.

The Court concluded at §95 that there was a prima facie case of indirect discrimination. This having been determined the Court did not go on to consider Whipple J’s approach to the claim based on Thlimmenos v Greece 31 EHRR 15 in any detail, though at §98 it suggested that it was flawed for similar reasons to her approach to indirect discrimination. The appeal ultimately failed on justification, the Court having reminded itself (per Lady Hale in R (SG) v Secretary of State for Work and Pensions) [2015] UKSC 16, [2015] 1 WLR 1449 §89) that “It is … the measure itself which has to be justified, rather than the fact that women are disproportionately affected by it”), and adopted the approach set out by Lord Reed in   Bank Mellat v HM Treasury (No.2) [2013] UKSC 39, [2015] 1 WLR 1449 §74. At §101 the Court ruled that disproportionate impact was not irrelevant to the justification test which “involves weighing the importance of the measure against its discriminatory impact”. At §104 the Court pointed out that, while correct at the time of Whipple J’s judgment, the “manifestly without reasonable approach” test had in fact come to be questioned in SC.

The Court went on to find however that the substance of Whipple J’s approach to justification had been correct: “Her assessment was appropriately nuanced, gave appropriate respect to the assessment of democratically accountable institutions, and recognised not only the need for caution before intervening in areas of social and economic policy but also that cogent justification was required for a measure having a differential impact on women… she applied a level of scrutiny appropriate to the circumstances of the case… We were not persuaded that there was any error of principle or any flaw in her reasoning. In those circumstances, it would be contrary to established authority for this Court to re-determine this issue which is fundamentally a question of proportionality” (§105). Further and in any event (§106) “even if it fell to this Court to revisit the judge’s assessment we would come to the same conclusion as she”. In explaining this conclusion the Court made reference to Lord Reed’s warning in SC §162 that “Since the principle of proportionality confers on the courts a very broad discretionary power, such cases present a risk of undue interference by the courts in the sphere of political choices. That risk can only be avoided if the courts apply the principle in a manner which respects the boundaries between legality and the political process” (§127) It stated that the decision under challenge “plainly came within the field of social and economic policy” and “was made by a democratically accountable government minister under powers vested in him by Parliament” and that the discrimination being indirect rather than direct “a less intense level of review may be appropriate”. While some of the respondent’s rationalisation was ex post facto and therefore required greater scrutiny (§128) there was no reason to conclude that they were other than bona fides and the Court accepted that they were “elaborations of arguments which were plainly under consideration when the scheme was being developed [a]nd …  received an appropriate level of scrutiny both at first instance and in this Court”. The court made reference at §129 to the “legitimate aims” served by the respondent’s approach: “effectiveness and speed of policy delivery, ease of verifying the figures supplied in order to reduce the risk of fraud (and other errors), the need to avoid perverse effects, and cost. In truth, it was the obvious means for the purpose, for all the reasons given by the respondents in their evidence: accounts submitted to HMRC over past years were the only verified information about self-employed earnings that was readily available”. Having concluded at §130 that the claimants’ “real complaint … is  … the failure to introduce a special modification to cater for the case of recent mothers whose profits in the measurement period will have been unrepresentative of their likely hypothetical no-Covid earnings in 2020:

“we have deliberately set out the history of the development of the scheme in some detail to illustrate the extreme if not unique circumstances in which it was devised. It is evident that it was not until the middle of March 2020 that the government appreciated the seriousness of the threat to lives and to the economy posed by Covid-19. By that time, the die was cast and the need for urgent action obvious. Time was of the essence. The scheme therefore had to be devised and constructed with the utmost speed. To obtain additional information in the case of recent mothers would have involved a complex exercise which in turn would have significantly delayed the implementation of the scheme. Furthermore, any further information would have had to be based on some form of self-reporting by the individual taxpayer of impact of maternity on their earnings in the measurement period: whatever method were adopted, its reliability and accuracy of which would be hard to verify. Perhaps, given time, it would have been possible to devise a modification that allowed recent mothers to submit verifiable information that could form the basis for a special assessment of their likely lost earnings. But in our judgment it would plainly not have been possible to do so in the limited time available without compromising the essential requirements of speed, simplicity and verifiability.

131 The judge also concluded that it would not have been possible to fashion a scheme that resolved the difficulties identified by the claimants without creating another range of hard cases and anomalies. On any view, devising a scheme that addressed the difficulties of those who took maternity leave without creating other perverse effects would plainly have taken time and resources. We do not accept that that difficulty can be ignored simply by reference to the special status of women on maternity leave.

In short, given the cardinal features required of the scheme – and above all speed and simplicity – the first respondent was in our view justified in introducing the scheme in a form which did not contain special provision for the position of recent mothers.”

Claimants; Jude Bunting, Clare Duffy and Donnchadh Greene, instructed by Leigh Day and Co

Defendants: Julian Milford QC and Rupert Paines, instructed by The Treasury Solicitor