The Latest from Cloisters

News, Publications, Policies, Events and Blogs

A Landmark Ruling on the Burden of Proof in Equal Pay: Armstrong & Others v Glasgow City Council [2017] CSIH 56

On Friday the Court of Session handed down an important judgment on the burden of proof and also the evidence needed in cases in equal pay claims where the claimants challenge their employer’s Job Evaluation Scheme (Armstrong & Others v Glasgow City Council [2017] CSIH 56). DAPHNE ROMNEY QC considers the judgment and its implications.


What happens if a woman’s work is evaluated by a job evaluation scheme which places her in a lower grade than a man, so that he earns more than she does, but where the scheme itself does not appear to be objective, analytical or fair? Equal pay law allows the scheme to be challenged but what evidence is required and who has the burden of proof? These were the questions for the Court of Session in a Cloisters-dominated case: the claimants were represented by Jason Galbraith-Marten in the ET, Robin Allen QC with Jonathan Mitchell QC in the EAT, and Jonathan Mitchell QC in this appeal to the Court of Session. Brian Napier QC acted for the respondent in this appeal.

Glasgow’s JES

In 1999 COSLA, the organisation representing Scottish local authorities, and the Scottish trade unions agreed the Single Status Agreement, (otherwise known as “the Red Book”). Like the English and Welsh Single Status Agreement (“the Green Book”) in 1997, the Scottish Single Status Agreement tried to make pay fairer for predominantly female roles which had been underpaid in relation to male manual workers. It also amalgamated the separate pay scales, terms and conditions for manual and administrative workers. Both Agreements set out a template for a new JES with agreed factors and marking levels. However no Council was obliged to apply that particular scheme.

With effect from April 2006, the respondent Council introduced a bespoke scheme, the Workplace Pay and Benefits Review (“WPBR”). Its JES fell into two parts, each leading to a separate payment. First there was a grade factor plan with eight factors, using a system of job profiles made up of composite jobs. These were assigned to a job family using an “allocation tool kit” which determined the employee’s grade, which in turn determined the employee’s core pay. There was also Working Context and Demands (“WCD”), an assessment of an employee’s five specific working conditions. WCD accumulated points for various factors, which did not affect the employee’s grade. However, points meant prizes, namely additional annual payments. As a result, the pay differential was maintained between the comparators (who had previously earned “productivity bonuses”) and the claimants.

The ET found that there was no precedent for this scheme; it had not been subject to peer review and did not follow the EOC advice. There were no job descriptions and there was no employee input, all of which was highly unusual and in marked contrast to the provisions of the Red Book. The ET also found that neither the WPBR nor the WCD contained all the factors necessary for a proper evaluation but that they could be read together to constitute a complete job evaluation. The question was therefore whether the methodology of two separate points systems leading to two separate payments complied with the statutory requirements for a job evaluation scheme or, alternatively, gave rise to reasonable suspicion of unsuitability.

The Law

In an equal pay claim, a woman must show that she does like work, work rated as equivalent or work of equal value to her male comparator (the same of course applies to a man and a female comparator). Where a JES has evaluated both the woman’s job and the comparator’s job and has ranked his job higher than hers, the woman is precluded from asserting that her work is of equal value and is bound by the results unless she can show that it does not comply with s. 65(4) Equality Act 2010 (“EqA”) and is not a thorough and objective analysis of her job and his. The alternative, even if s. 65(4) is satisfied, is for her to adduce evidence such that an ET can conclude that there are reasonable grounds to suspect that the JES is unreliable (s. 131(6) EqA).

Under the Equal Pay Act 1970 (“EPA”), which was replaced by the EqA, s. 1(5) mirrored s. 65(4) EqA and s. 2A(2A) was broadly similar to s. 131(6) (EqA), save that it referred to a JES being “otherwise unsuitable” to be relied upon rather than “unreliable”, the current formulation. This case was decided under the applicable provisions of the EPA.

In Bromley v Quick [1988] I.C.R. 623, Dillon LJ at p. 633 approved the EAT’s decision in Eaton v Nuttall [1977] I.C.R. 272 at 277, namely:

“a study satisfying the test of being thorough in analysis and capable of impartial application. It should be possible by applying the study to arrive at the position of a particular employee at a particular point in a particular salary grade without taking other matters into account except those unconnected with the nature of the work”.

Burton J expressed similar views in Diageo v Thomson [EATS/0064/03]. The scheme must also be free of sex taint – see Rummler v Dato-Druck G.m.b.H [1987] I.C.R. 774, where the ECJ held that factors likely to favour men (in that case, physical effort) had to be balanced with factors which favoured women.

There have been no previous appellate decisions on the suitability of a JES but ETs have considered this issue. The NHS job evaluation scheme, Agenda for Change, was upheld in Hartley v Northumbria Healthcare NHS Foundation Trust ET/2507033/2007. The ET said at para. 584.4,

“The question for determination will be whether the Tribunal, having considered the evidence on both sides, is left with reasonable grounds for suspicion, even if there is not sufficient evidence to make a positive finding of discrimination or unsuitability.”

On the other hand, ETs have found the JES to be unsuitable in the cases of Russell v South Lanarkshire ETS/107667/05 (which was another bespoke scheme), Brennan v Sunderland ETS/2503297/06 and Surtees v Middlesbrough ETS/2501390/03. The ET in Russell at paras. 301/302 made similar observations to those in Hartley about the requisite burden of proof.

The appeal

The ET held that the JES did comply with s. 1(5), chiefly because the claimants had failed to adduce any evidence (and in particular any expert evidence) to show why it did not. It said: “the effect of the claimants’ submissions is to seek to create a lacuna – the absence of a mechanism or rule for bringing the grade and WCD scores together – rather than to work with the methodology as it is”. In the absence of such evidence, the ET drew the inference that the combination of the two marks was legitimate, despite its own conclusion that the scheme was untested. The Court of Session was scathing, noting that the ET had erred on the burden of proof. It was for the respondent to show that the scheme satisfied s. 1(5), rather than for the claimants to show that it did not. The Court of Session said in para. 18 of the judgment: “the respondent failed to discharge the burden of proving that its two schemes together produced a rating and constituted a valid JES that satisfied the requirements of section 1(5).” It was not for the claimants or for the ET to fill the gap, but for the respondent (para. 22).

The Court of Session went further. Normally, for a respondent to show compliance with s. 1(5), it was insufficient just to produce the scheme “and leave the claimants to pick holes in it…. Discharging the burden of proof requires the positive obligation of leading evidence to document the scheme against the factors” (para. 41). A JES had to be “rigorously tested” against relevant factors and subject to ”rigorous analysis”. It follows that a respondent must adduce a positive case. It rejected the respondent’s suggestion that it was enough to show a prima facie case. The Court of Session also stressed that the ET had made three important findings:

  1. this scheme was “novel and untested”;
  2. there was no expert evidence to support its design; and
  3. there was no way to aggregate the two values from the parallel exercises.

In a scheme with such unusual features (para. 43) the respondent was obliged to show that it met the requirements of s. 1(5).

Expert evidence

Throughout the case, the ET had bemoaned the absence of an independent expert witness who could assist with the design of the JES and claimed it was unable to decide the issue. Nevertheless, it then made an assessment despite the lack of expert evidence it had complained of. At para. 54, the Court of Session said:

“It was not part of the function of the ET to speculate as to whether aspects of the JES might be made to work in such a way as to render them compliant. If there was a lacuna in the methodology of the JES, it was not part of the ET’s function to try to fill that lacuna. If the Tribunal could not be satisfied on the basis of the evidence led before it that the methodology of the JES was justified and its analysis thorough, the ET required to find that it was not a valid job evaluation as defined in section 1(5) of EQP.”


The Court of Session noted drily that the ET’s statement that it was not in a position to resolve the issue was not a case of the specialist ET asserting knowledge, which is what ETs are supposed to be,  but of an ET expressly disavowing expertise (para. 55). The absence of expert evidence meant that the respondent had failed to discharge the burden and the ET should have found accordingly. 

s. 2A(2A)

That disposed of the appeal but the Court of Session went on to consider the claimants’ alternative ground of appeal under s. 2A(2A). It was accepted that unlike s. 1(5), it was for the claimants to show that there were grounds for reasonable suspicion and they had the burden of proof. However, it emphasised (para. 22) that whilst the burden was on the claimants to adduce that evidence, it was for the ET to make the assessment. Again, the Court of Session stated that the ET had got the burden of proof wrong. It was not for the claimants to produce “cogent evidence”  that the JES was unsuitable to be relied upon; previous decisions in other tribunals had held that the burden of proof on the claimants was far less onerous, as stated in Russell and Hartley.

The claimants submitted that each of the three specific findings set out above gave rise to a reasonable suspicion whilst the Council submitted that all that the EPA required was a mechanism for comparing the two jobs. The Court of Session held that there was evidence which suggested that the JES was unsuitable and that the ET should have come to that conclusion.

On the issue of expert evidence, the Court of Session held that although claimants could lead expert evidence, they were not obliged to do so since they were required only to show reasonable grounds for suspicion.


This decision has made it easier for claimants to challenge a JES which prevents them from claiming equal value with a comparator. The hearings in Surtees,  Hartley,  Russell and Brennan have been very lengthy and expensive and have delayed the equal value process for claimants. The question of whether to adduce expert evidence, and if so from whom, has also produced confusion and delay. The Court of Session has made the need for expert evidence redundant – it is for the respondent  to prove that the JES complies with s. 1(5) and to explain what is, after all, within its own knowledge. Conversely, whilst the claimants have to point to matters which raise a reasonable suspicion in the ET’s mind, the matters need only to be shown, not actually proved to be the case, which eases the burden considerably.  Above all, it is for the employer introducing a job evaluation scheme to ensure that it is objective, analytical and fair and that there is an audit trail showing how decisions were reached- and why.

Not all in the family: migrant domestic worker who...
A radical reconsideration of the burden of proof: ...