powered by Zehuti
Appeal against the dismissal of Unison's judicial review application against the introduction of tribunal fees. Appeal dismissed.
Unison brought judicial review proceedings challenging the Government's introduction of tribunal fees. The challenge was rejected and Unison appealed.
The appeal was dismissed, Lord Justice Underhill saying:
"I have a strong suspicion that so large a decline [in claims] is unlikely to be accounted for entirely by cases of "won't pay" and that it must also reflect at least some cases of "can't pay"; and I have accordingly been tempted by Ms Monaghan's submission that the figures speak for themselves. But in the end I do not think that that is legitimate. The truth is that, looked at coolly, there is simply no safe basis for an untutored intuition about claimant behaviour or therefore for an inference that the decline cannot consist entirely of cases where potential claimants could realistically have afforded to bring proceedings but have made a choice not to. But in fact the difficulty goes further than that. Even if it really were an irresistible inference from the decline in claims that at least some potential claimants could not realistically afford the fees, there is still no basis for forming any reliable view about the numbers of such cases, or how typical they may be; and, for reasons which will appear, that is an important matter. In my view the case based on the overall decline in claims cannot succeed by itself. It needs to be accompanied by evidence of the actual affordability of the fees in the financial circumstances of (typical) individuals. Only evidence of this character will enable the Court to reach a reliable conclusion that that the fees payable under the Order will indeed be realistically unaffordable in some cases."
Neutral Citation Number:  EWCA Civ 935
Case No: C1/2015/0059
IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
Lord Justice Moses and Mr Justice Irwin
 EWHC 218 (Admin)
AND ON APPEAL FROM THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
Lord Justice Elias and Mr Justice Foskett
 EWHC 4198 (Admin)
Royal Courts of Justice
Strand, London, WC2A 2LL
LORD JUSTICE MOORE-BICK
(Vice President of the Court of Appeal, Civil Division)
LORD JUSTICE DAVIS
LORD JUSTICE UNDERHILL
- - - - - - - - - - - - - - - - - - - - -
THE QUEEN (ON THE APPLICATION OF UNISON) (Claimant/Appellant)
- and –
THE LORD CHANCELLOR (Defendant/Respondent)
- and -
EQUALITY AND HUMAN RIGHTS COMMISSION (Intervener)
- - - - - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - - - - - - - -
Ms Karon Monaghan QC and Mr Mathew Purchase (instructed by Unison Legal Services) for the Appellant
Mr David Barr QC and Ms Susan Chan (instructed by Government Legal Department) for the Respondent
Mr Michael Ford QC (instructed by Equality and Human Rights Commission) for the Intervener (written submissions only)
Hearing dates: 16 & 17 June 2015
- - - - - - - - - - - - - - - - - - - - -
Lord Justice Underhill :
1. Until 29 July 2013 claimants could pursue proceedings in an employment tribunal (an "ET") without paying any fee to Her Majesty's Courts and Tribunals Service ("HMCTS"); likewise appellants to the Employment Appeal Tribunal ("the EAT"). By the Employment Tribunal and the Employment Appeal Tribunal Fees Order 2013 (SI 2013/1893) ("the Fees Order"), made by the Lord Chancellor under statutory powers, that has ceased to be the case. Under the Order fees are payable by the claimant or appellant on the commencement of a claim or appeal and also in advance of the final hearing, unless they are entitled to a remission on account of limited means: I give fuller details below.
2. On 1 July 2013 the trade union Unison commenced judicial review proceedings challenging the Lord Chancellor's decision to make the Fees Order, which had been laid before Parliament but not yet made. I need not at this stage set out the grounds of challenge in full, but a central ground was that the fees were set at such a level, and the remission criteria were so restricted, that many claimants would be unable to afford to bring a claim in the tribunals. I will refer to these proceedings as Unison 1.
3. Unison sought a stay on the making of the Fees Order pending the decision on its challenge. In his response the Lord Chancellor volunteered an undertaking "that, if he is allowed to make the Order and if, following exhaustion of appeal rights, it is found to be unlawful, he will reimburse all fees that have been paid". That undertaking was apparently tendered to Lewis J at a hearing on 29 July 2013, though it was not (as it should have been) incorporated in the recital to his order; and no stay was ordered. The Order came into force on that day.
4. The claim in Unison 1 was heard by a Divisional Court comprising Moses LJ and Irwin J over two days in mid-October 2013 and on a third day in November. The Equality and Human Rights Commission ("the Commission") intervened in support of the challenge. By a judgment handed down on 7 February 2014 ( EWHC 218 (Admin),  ICR 498) the claim was dismissed. A principal reason why the challenge failed was that the Court believed that it was premature, because it was wrong to reach a decision on the impact of the Fees Order on the basis of predictions rather than wait until it was possible to see what its effect had been in practice. Consistently with that reasoning, the Court refused permission to appeal, although it was acknowledged that the claim raised some arguable issues, on the basis that the right course was for Unison to launch further proceedings, if it chose, once better information was available.
5. Unison did not, and does not, accept that the challenge in those proceedings was premature. It applied to this Court for permission to appeal. On 19 May 2014 Sedley LJ gave limited permission and also made an order permitting Unison to adduce evidence about the reduction in the volume of claims to the ET and appeals to the EAT in the period since fees had been introduced.
6. Unison renewed its application orally in relation to the grounds on which permission had not been given; and the Lord Chancellor in his turn applied to have the order for the admission of further evidence set aside. However, when those applications came before myself and Sharp LJ in September 2014 Unison had announced that it was about to issue further proceedings, and it was agreed that the applications should be stood over for the time being.
7. The further proceedings ("Unison 2") were filed on 23 September 2014. The grounds of challenge overlap with those in Unison 1 but Unison was able to rely on evidence about the actual impact of the fees in the period of more than a year since they were introduced.
8. Unison 2 was heard in October 2014 by a Divisional Court comprising Elias LJ and Foskett J. The Commission again appeared as an intervener. By a judgment handed down on 17 December 2014 ( EWHC 4198 (Admin),  ICR 390) the claim was dismissed.
9. The Divisional Court itself gave permission to appeal against that decision. The parties then sought to revive the applications in Unison 1 to which I have referred at para. 6 above. On 31 March 2015 Burnett LJ and I gave permission to appeal as regards the remaining grounds and directed that both appeals be heard together. We stood over to the hearing of the substantive appeals the Lord Chancellor's application to set aside the order of Sedley LJ permitting the introduction of fresh evidence.
10. Accordingly what is before us are the appeals against the decisions in both Unison 1 and Unison 2. Ms Karon Monaghan QC and Mr Mathew Purchase appear for Unison, as they did at first instance in both cases. Mr David Barr QC and Ms Susan Chan appear for the Lord Chancellor; Ms Chan appeared on her own at first instance in both cases. The Commission has again been permitted to intervene, but on this occasion it has done so only in the form of written submissions settled by Mr Michael Ford QC.
THE FEES ORDER
The Statutory Basis
11. Section 42 (1) of the Tribunals, Courts and Enforcement Act 2007 confers power on the Lord Chancellor by order to prescribe fees payable in respect of "anything dealt with by" the First-tier and Upper Tribunals and also by an "added tribunal". By section 43 (3) the Lord Chancellor was given powers to specify "added tribunals": the ET and the EAT were so specified by the Added Tribunals (Employment Tribunals and Employment Appeal Tribunal) Order 2013 (SI 2013/1892).
The Introduction of the Order
12. In January 2011 the Government announced that it was its view that in principle users of the ET and the EAT should be asked to contribute towards the costs of running them by the payment of fees. Detailed proposals were advanced for consultation in December 2011. A response to consultation was issued in July 2012. On 25 April 2013 a draft of the eventual Fees Order was laid before Parliament. It was debated and approved by both Houses under the affirmative resolution procedure. It was made on 28 July 2013 and, as I have said, came into force on the following day.
13. Article 3 of the Fees Order makes general provision for fees to be payable in respect of any claim presented to an ET or of an appeal to the EAT. Fees in the ET are governed by Part 2 and fees in the EAT by Part 3.
14. So far as fees in the ET are concerned, article 4 defines the "charging occasions" on which fees will become payable by claimants. There are two basic charging occasions – when a claim is presented ("the issue fee"), and when notice is given of the listing of a final hearing ("the hearing fee"). Fees are also chargeable on the making of various kinds of application. The amounts chargeable are specified in tables set out in Schedule 2. So far as the issue fee and hearing fee are concerned, these vary depending on whether the claim is brought by a single claimant or a group and on whether the claim is classified as "type A" or "type B". Table 2 defines 61 specific types of claim as type A; type B are the rest. Broadly speaking, type A claims are those which the Lord Chancellor regards as typically the more straightforward: I annex a copy of the table. Unfair dismissal claims and discrimination claims do not appear in the table and accordingly fall into type B1. The issue fee and hearing fee for a single claimant bringing a type A claim are £160 and £230 respectively (thus £390 in total); for a type B claim the figures are £250 and £950 (£1,200 in total). For multiple-claimant claims the figures depend on the numbers in the group. I need not give all the variables. For the smallest group (up to ten claimants) the totals are £780 for type A claims and £2,400 for type B claims; for the largest (over 200) they are £2,340 for type A and £7,200 for type B. I need not give the figures for applications.
15. In the EAT £400 is payable when the appeal is issued and £1,200 when the EAT directs that the appeal proceed to an oral hearing. There is no distinction between different "types" of appeal or between single and group appeals.
16. Article 17 enacts Schedule 3, which contains the provisions relating to fee remission. These reproduce the regime for remission of fees in the ordinary courts. That regime changed soon after the Fees Order came into force, with the result that a new Schedule 3 was substituted with effect from 7 October 2013 (see the Courts and Tribunals Fee Remissions Order 2013 ("the Remissions Order")). For present purposes we are primarily concerned with that later version, but I shall need to say a little about its short-lived predecessor: see paras. 25-26 below.
17. Disposable capital. Claimants and appellants2 will not be entitled to fee remission unless they satisfy "the disposable capital test", i.e. that their disposable capital is less than a specified amount: this is the subject of paras. 3-10 of the Schedule. The amount varies according to the amount of the fee. Where (as is the case for all the fees with which we are concerned for single claimants, except for the second stage in the EAT) the fee is no more than £1,000, the amount of capital specified is £3,000; the amount applicable to a fee of £1,200 is £4,000. There are elaborate provisions defining "disposable capital" but I need not summarise them here.
18. Income. If a claimant satisfies the disposable capital test he or she will be entitled to full fee remission if their gross monthly income is below a specified amount: this is the subject of paras. 11-13 of the Schedule. The amount varies depending on whether the party is single and whether he or she has children. The figures are as follows:
The figures go up by £245 for each additional child. (The annualised equivalents of the figures at the top and bottom of that range – that is, ignoring additional children – are £13,020 and £20,820.) For every £10 of gross monthly income above the specified amount the claimant must pay £5 towards the fee. Thus, by way of example, a single claimant with no children bringing a type B claim would have to pay a full issue fee once their monthly income exceeded £1,545 (annual equivalent £18,540) and a full hearing fee once it exceeded £2,985 (£35,820).
19. Partner's means. By para. 14 of the schedule, for the purpose of the calculation both of a claimant's disposable capital and of their monthly income the capital and income of a partner is taken into account. The provision is disapplied where the partner has a contrary interest.
20. Procedure for application. An application for remission must be made at the time when the fee would otherwise be payable. The potential obligation to pay the fee is, in effect, suspended while the application is processed. It is determined by a fees remission office within HMCTS for which the Lord Chancellor is ultimately responsible. There is no judicial involvement in the process.
21. Paragraph 16 of Schedule 3 reads as follows:
"A fee specified in this Order may be remitted where the Lord Chancellor is satisfied that there are exceptional circumstances which justify doing so."
22. Guidance published by HMCTS, entitled Court and Tribunal Fees – Do I Have to Pay Them?, contains a section headed "Exceptional Circumstances". This reads:
"If your fee remission application is refused, you may not have to pay a court or tribunal fee if you have suffered an unexpected event, that has seriously affected your ability to pay a court or tribunal fee. If you can prove this circumstance is exceptional, for example, letter or notices threatening legal action due to non payment of bills, the Delivery Manager has the power to grant a full or part remission.
The Delivery Manager is the only person who can make this decision and it is based on the information given to the court or tribunal at the time that the court or tribunal fee is due. The Delivery Manager's decision is final and cannot be appealed.
Examples of exceptional circumstances may be when:
* payment of a fee would mean non-payment of an essential service or utility bill (for example, electricity or gas) that is likely to lead to the service being cut off;
* payment of a fee would mean non-payment of rent or mortgage amounts that are overdue, which could lead to you being made homeless;
* you have personal responsibility for caring for a dependent adult and that care can only be paid for from your own resources;
* you have suffered unexpected and sudden personal and financial loss or expense due to the death of a close family member or dependent relative; or you cannot pay the fee due to uninsured loss or damage to personal belongings as a result of fire, flood, theft or criminal damage."
23. We were also shown internal HMCTS guidance relating to the exercise of this discretion. This broadly reflects the published guidance, but I should note the following points:
(1) The discretion is described as applying where payment of the fee "would, owing to the exceptional circumstances of the particular case, involve undue hardship (financial or otherwise) to the applicant". That does not precisely reflect the language of paragraph 16, which does not prescribe a criterion of "undue hardship", but here is a fairly detailed discussion of what might constitute "financial or other hardship". Reference is made to whether payment of the fee "would seriously impact on [the applicant's] day-to-day life" and the same examples are given as in the published guidance.
(2) It is said that "remissions granted under this criterion will be rare".
(3) Contrary to what is said in the published guidance, it is made clear that an adverse decision by a Delivery Manager can be appealed to the Operations Manager. Mr Barr told us the published guidance would be altered to reflect the internal guidance.
24. Neither of those documents was put before either of the Courts below, and Elias LJ observed, at para. 61 of his judgment, that "we have heard little about this discretion or when it may be exercised". We ourselves have seen no evidence about how many applications have been made under paragraph 16, let alone their outcome.
The earlier remission regime
25. The remission regime in force between 29 July and 6 October 2013 allowed full remission of fees for claimants on certain specified benefits or whose gross annual income was below specified limits which varied according to whether they were single or had children. They were also entitled to full remission if their disposable monthly income was £50 or less, and to partial remission on a tapering basis above that figure: I need not give the details. There were elaborate provisions governing the deductions to be made from gross monthly income in order to arrive at disposable income: these were tax, national insurance, cost of accommodation, child care costs, child maintenance payments, sums payable pursuant to a court order and, finally, a fixed allowance intended to represent "cost of living expenses" (which, again, varied according to whether the applicant was single or had children). Those amounts were plainly intended to represent essential expenditure, and I will refer to them as such. There was also a disposable capital test, but I can ignore that for present purposes.
26. It will be seen that a principal difference between the two schemes is that the original version requires, in cases above a minimum annual income level, a case-by-case assessment of monthly disposable income, while the current version proceeds by reference only to gross income, evidently on the basis that at the levels specified the claimant will normally have sufficient disposable income, i.e. income over and above what is required for essential living expenses: the only concession to differences in expenditure is the different income levels specified for applicants who have partners or children.
Effect of Non-Payment
27. Both the Employment Tribunal Rules of Procedure (which are Schedule 1 to the Employment Tribunals (Constitution and Rules of Procedure) Regulations 2013) and the Employment Appeal Tribunal Rules 1993 provide, in summary, that a claim or appeal cannot proceed unless accompanied by a fee or a remission application: see rules 11 and 14 of the former and rule 17A of the latter.
Recovery of Fees by Successful Parties
28. Rule 76 (4) of the Employment Tribunal Rules and rule 34A (2A) of the Employment Appeal Tribunal Rules give the ET and the EAT respectively discretion to make an order that one party pay to the other the amount of any fees paid under the Fees Order. A series of decisions in the EAT have held, unsurprisingly, that such an order should normally be made in favour of a successful party, though it will not be appropriate in every case (e.g. where their success has been only partial) – see Portnykh v Nomura International plc  UKEAT 448/13, Horizon Security Services Ltd v Ndeze  UKEAT 71/14,  IRLR 854, and Look Ahead Housing and Care Ltd v Chetty  UKEAT 37/14,  ICR 375. Those cases were all concerned with costs in the EAT but no doubt the same principles apply in the ET.
29. Unison's original challenge to the Fees Order was on four bases:
(1) Breach of the Principle of Effectiveness. Many of the types of claim over which the ET and the EAT (to which I will refer compendiously as "the Tribunals", except where it is necessary to distinguish) have jurisdiction are derived from EU law.3 It is Unison's case that the regime introduced by the Fees Order breaches the EU "principle of effectiveness" by making it impossible in practice, or excessively difficult, for claimants to enforce those rights. This challenge was argued in the Divisional Court in both Unison 1 and Unison 2.
(2) Breach of the Principle of Equivalence. It was argued in Unison 1 that the fees payable in respect of claims based on EU-derived rights meant that such claims were subject to a less favourable regime than those based on domestic rights. The point was not re-argued in Unison 2, because it was not affected by the evidence of the impact of the fee regime in its first year of operation.
(3) Indirect Discrimination. The Fees Order is said to be indirectly discriminatory against claimants with particular protected characteristics. This challenge was argued in both Courts below.
(4) The Public Sector Equality Duty. It was argued in Unison 1 that in deciding to make the Fees Order the Lord Chancellor acted in breach of the duty imposed by section 149 of the Equality Act 2010. The point was not re-argued in Unison 2 for the same reason as at (2) above.
30. In its grounds of appeal in the two appeals Unison challenged the rejection of its case by the Divisional Court under each of those heads; but in the course of the hearing before us Ms Monaghan abandoned the challenge based on the principle of equivalence. It also, as a separate ground, challenged the characterisation of its claim in Unison 1 as premature. That was said to be potentially relevant not only to the issue of costs but also to whether the Lord Chancellor was bound by the undertaking which he gave in Unison 1 – see para. 3 above – which was not repeated in the context of Unison 2; but Mr Barr took the wind out of those particular sails by acknowledging at the start of the hearing before us that the Lord Chancellor could not in practice resist claims for repayment of fees paid under an invalid order.
31. I will take in turn the three surviving grounds of challenge and finally address the question of prematurity.
(A) BREACH OF THE EFFECTIVENESS PRINCIPLE
32. It is a principle of EU law, now embodied in article 47 of the EU Charter, that persons who claim that their rights under EU law have been breached must have access to an effective remedy for that breach: otherwise the rights in question would be illusory. It is hardly necessary to cite authority for so basic a proposition, but we were referred in particular to Johnston v Chief Constable of Royal Ulster Constabulary (C-222/84),  QB 129, at para. 17 (p. 147), and to Levez v T. H. Jennings (Harlow Pools) Ltd (C-326/96),  ICR 521, which seems to be the source of the label "the effectiveness principle" (see para. 18 (p. 541H)). But in truth that is only a reflection, as the ECJ has itself said (see, e.g., Johnston, at para. 18 (p. 147 G-H)), of a principle common to the legal systems of all member states4 and also in the case-law of the European Court of Human Rights ("the ECtHR") relating to article 6 of the European Convention of Human Rights. We were accordingly referred indifferently to authorities from both the ECJ/CJEU and the ECtHR; and it is of course now established that the jurisprudence of the Strasbourg Court is authoritative in Luxembourg – see, e.g., Criminal Proceedings against Pupino  ECR I-5828, paras. 58-59, and article 52 (3) of the European Charter of Fundamental Rights.
33. A central aspect of that principle is that potential claimants must have effective access to a court or tribunal (I will use "court" as a shorthand) which can afford them such a remedy. "Effective" in this context connotes that the ability of the claimant to invoke the assistance of the court must be real rather than merely theoretical. That point is made, for example, by the ECtHR in Airey v Ireland (1979) 2 EHRR 305, which concerned access to legal aid. The Court said, at para. 57:
"... The Convention is intended to guarantee not rights that are theoretical or illusory but rights that are practical and effective ... This is particularly so of the right of access to the courts in view of the prominent place held in a democratic society by the right to a fair trial ... It must therefore be ascertained whether Mrs. Airey's appearance before the High Court without the assistance of a lawyer would be effective, in the sense of whether she would be able to present her case properly and satisfactorily."
34. Unison's case is that the regime under the Fees Order – that is, the combination of the level of the fees and the criteria for remission – is such that many claimants are unable to afford to bring claims in the Tribunals and are accordingly denied effective access to justice; and that the Order is accordingly unlawful. That case is, as I have said, framed by reference to the EU principle of effectiveness and thus with reference to claims in relation to EU-derived rights; but the equivalent reasoning by reference to article 6 of the Convention would appear to mean that it could apply to any claim within the jurisdiction of the Tribunals5.
35. Two general questions need to be considered before I turn to the evidence relied on by Unison in support of its case.
(1) THE TEST OF AFFORDABILITY
36. We heard a fair amount of argument about the criterion to be applied in deciding whether a claimant could afford a particular level of fee. It was accepted that the starting-point was the decision of the ECJ in Levez. That was not a case about court fees – it concerned the then two-year limitation period applying to claims under the Equal Pay Act 1970 – but in it the Court referred to a number of previous authorities about procedural conditions affecting the entitlement to bring claims to vindicate EU-derived rights and held (see para. 18, at p. 541 G-H) that such conditions were unobjectionable
"… provided … that [they] … do not render virtually impossible or excessively difficult the exercise of rights conferred by Community law."
Broadly the same principle is expressed in numerous subsequent cases. However, in some of them6 the formulation "impossible in practice" is used, and although the two phrases are plainly meant to mean the same thing I think that this is to be preferred, because it more clearly conveys that a realistic approach to assessing "impossibility" is required. It is also more in keeping with the alternative phrase "excessively difficult": that, again, must be intended to mean the same as "virtually impossible/impossible in practice" – otherwise the latter phrase would render the former redundant – but the word "excessively" recognises that judgments of degree are required.
37. Ms Monaghan and Mr Ford submitted that the later case-law of the CJEU showed that the concepts of "virtual impossibility" and "excessive difficulty" were not in practice treated as setting up as high a hurdle as the language might suggest, and indeed that the essential test was one of proportionality. They relied in particular on the decisions in Impact v Minister for Agriculture and Food (C-268-06),  ECR I-2483, Alassini v Telecom Italia SPA (C-317-320/08),  3 CMLR 17, and Sociedad Duarte Hueros v Autociba SA (C-32/12)  1 CMLR 53. These cases are discussed in some detail by Elias LJ at paras. 29-32 and 40-43 of his judgment in Unison 2, and I would not dissent from his illuminating analysis; but I find them unhelpful on the more straightforward basis that they are not concerned with the question of court fees, and it is difficult to extract from them any clear statement of general principle which either modifies the formulation in Levez or casts light on how it should be applied in a case like the present. I find more assistance from three decisions of the ECtHR in which the issue of whether the imposition of court fees unlawfully hindered access to justice arose directly – Kreuz v Poland (no. 1) (28249/95), (2001) 11 BHRC 456; Podbielski and PPU Polpure v Poland (39199/98),  ECHR 543; and Apostol v Georgia  ECHR 999. We were in fact referred directly only to Podbielski, no doubt because it contains the fullest discussion of the correct approach, but Kreuz is referred to by the Court in Unison 1 and Apostol by Elias LJ in Unison 2.7
38. In Podbielski an insolvent company and its owner were involved in commercial litigation involving very substantial sums. Under the Polish system court fees vary according to the amounts in issue, although the court has power to order full or partial exemption. It was claimed that the company had been required to pay a fee of PLN 10,000 which it could not afford. The Court expressed the applicable principles at paras. 61-64 of its judgment as follows (omitting some irrelevant matter):
"61. Article 6.1 secures to everyone the right to have any claim relating to his civil rights and obligations brought before a court or tribunal. In this way, that provision embodies the "right to a court", of which the right of access, that is the right to institute proceedings before a court in civil matters, constitutes one aspect only; however, it is an aspect that makes it in fact possible to benefit from the further guarantees laid down in paragraph 1 of Article 6 (see, among many other authorities, Golder v. United Kingdom, judgment of 21 January 1975, Series A no. 18, pp. 16-18, paras. 34 in fine and 35-36, and Kreuz (no. 1) ... (paras. 52 et seq.)
62. The "right to a court" is not absolute. It may be subject to limitations permitted by implication because the right of access by its very nature calls for regulation by the State. Guaranteeing to litigants an effective right of access to courts for the determination of their "civil rights and obligations", Article 6 § 1 leaves to the State a free choice of the means to be used towards this end but, while the Contracting States enjoy a certain margin of appreciation in that respect, the ultimate decision as to the observance of the Convention's requirements rests with the Court.
63. The Court has accepted that in some cases, especially where the limitations in question related to the conditions of admissibility of an appeal, or where the interests of justice required that the applicant, in connection with his appeal, provide security for costs to be incurred by the other party to the proceedings, various limitations, including financial ones, may be placed on his or her access to a "court" or "tribunal". However, such limitations must pursue a legitimate aim and there must be a reasonable relationship of proportionality between the means employed and the legitimate aim sought to be achieved ...
64. The requirement to pay fees to civil courts in connection with claims, or appeals, they are asked to determine cannot be regarded as a restriction on the right of access to a court that is incompatible per se with Article 6.1 of the Convention. However, the amount of the fees assessed in the light of the particular circumstances of a given case, including the applicant's ability to pay them, and the phase of the proceedings at which that restriction has been imposed are factors which are material in determining whether or not a person enjoyed his right of access and had "a ... hearing by [a] tribunal" (see Kreuz (no. 1) and Tolstoy-Miloslavsky v United Kingdom [(1995) 20 EHRR 442)]."
It went on to apply those principles to the facts of the case. It observed that the situation was not comparable to the case of security for costs (which was the subject of Tolstoy-Miloslavsky), because court fees had nothing to do with the protection of the other party or the merits of the claim but were purely concerned to raise revenue for the state (see para. 66). It said that in such a case whether they constituted a restriction "should be subjected to a particularly rigorous scrutiny from the point of view of the interests of justice" (para. 65). It then pointed out that the company had by that date had all its assets attached by creditors and all its bank accounts frozen, so that it could plainly pay nothing from its own resources (see para. 67), and the real questions were whether the Polish court was entitled to take the view that it should have put money aside in the past to fund the litigation and whether it was material that Mr Podbielski had been able to pay a similar court fee two years later. After discussing those questions it said, at para. 69:
"In the circumstances and having regard to the prominent place held by the right to a court in a democratic society, the Court considers that the judicial authorities failed to secure a proper balance between, on the one hand, the interest of the State in collecting court fees for dealing with claims and, on the other hand, the interest of the applicant in vindicating his claim through the courts."
Its overall conclusion, at para. 70, reads:
"The Court therefore concludes that the imposition of the court fees on the applicant constituted a disproportionate restriction on his right of access to a court. There has accordingly been a violation of Article 6.1 of the Convention."
39. Although the reasoning and language in Kreuz are very similar to Podbielski, it is worth noting that in its conclusion (at para. 66) the Court observed:
"The fee required from the applicant for proceeding with his action was excessive. It resulted in his desisting from his claim and in his case never being heard by a court. That, in the Court's opinion, impaired the very essence of his right of access."
40. In my view Podbielski clearly establishes:
(a) (unsurprisingly) that it is not objectionable in principle for the state to charge a fee for access to the courts;
(b) that there should be "a proper balance" between the right to charge such a fee and the right of a claimant to bring a claim before the court; and
(c) that the balance will not be properly struck if the fee is "disproportionate".
However, there is no explicit discussion of what judging "proportionality" involves in this context. Reference to the formulation of the classic proportionality test does not help, even if, which I rather doubt, that was what the Court had in mind. It accepted that seeking a contribution from litigants to the cost of the court system was legitimate: the question is at what level the fee ceases to bear, as it put it at para. 62, "a reasonable relationship" to that legitimate aim – or, as it is more bluntly put in Kreuz, becomes "excessive". That question did not require discussion in any of these three cases because on the facts the fee was plainly unaffordable. In Kreuz it represented "a sum equal to the average annual salary in Poland at that time" (see para. 62 of the judgment) and the claimant had no income or savings (loc. cit.) and was, on the evidence, simply unable to pay it (para. 65). In Podbielski the company was insolvent. In Apostol the amount of the fee (which was in fact for the enforcement of a judgment) seems to have been only about the equivalent of €115; but the claimant's monthly income was a pension equivalent to €19 per month, and it was accepted that he simply could not pay at all.
41. Thus although some help is to be got from the case-law it does not provide any clear criterion for identifying at what level a particular court fee becomes "excessive" or "disproportionate". It is necessary to go back to the underlying principle. That is that a claimant must not be denied access to a court, and thus access to justice; and that such access should be practical and not merely theoretical. In my view it follows that the basic question is whether the fee payable is such that the claimant cannot realistically afford to pay it. If that seems a trite conclusion to so elaborate a discussion I must apologise. But it does add something. It means that the focus is squarely on what the claimant can afford to pay (rather than, for example, considerations of the value of the claim or the cost of the service), and it also emphasises that the approach must be realistic. The criterion of whether the claimant can realistically afford to pay the fee is consistent with the well-established test in Levez, albeit that that was not concerned specifically with court fees, provided due weight is given to the phrase "impossible in practice". As I read it, this was the approach followed by both Divisional Courts.
42. I would add a few further observations about that formulation.
43. First, what claimants can afford depends not only on their income but equally on their expenditure. Once essential living expenses are met (though I accept that what is or is not "essential" itself involves questions of judgment ), any remaining income may be treated as disposable, and whether claimants can afford to pay a court fee will depend on what else they choose to spend their income on. It follows that a claimant cannot be said to be unable to afford to pay a court fee simply because the choice to forgo other expenditure may be difficult. The test is whether any such difficulty is "excessive", i.e. such as to make payment of the fee impossible "in practice". Elias LJ made this point at para. 61 of his judgment in Unison 2, as follows:
"The question many potential claimants have to ask themselves is how to prioritise their spending: what priority should they give to paying the fees in a possible legal claim as against many competing and pressing demands on their finances? And at what point can the court say that there is in substance no choice at all? Although Ms Monaghan would not accept that this is the task facing the court, it seems to me that in essence that is precisely what the court has to do. In that context, as Moses LJ said in the first Unison challenge, it is not enough that the fees place a burden on those with limited means. The question is not whether it is difficult for someone to be able to pay - there must be many claimants in that position - it is whether it is virtually impossible or excessively difficult for them to do so. Moreover, the other factors which I have identified as potentially inhibiting a worker from pursuing a claim may reinforce the conclusion that the risks inherent in litigation are not worth taking. These factors engender a cautious approach to litigation but do not compel the inference that it would be impossible in practice for some of these claimants to litigate. "
44. Secondly, Ms Monaghan and Mr Ford emphasised that some of the rights which claimants seek to vindicate through the Tribunals are recognised in the jurisprudence of the EU as fundamental rights – principally, rights not to be discriminated against on the grounds identified in the various applicable Directives (being Council Directives 2000/43/EC (race), 2000/78/EC (all other protected characteristics except sex), and 2006/54/EC (sex)). I do not believe that in this particular context that makes a difference. In my view the right to access to a court to enforce a person's legal rights is itself a fundamental right, protected indeed by article 6 of the Convention; and I do not think that it is correct in principle to take a different approach to the affordability of access to justice depending on the nature of the claim. (I would not in any event accept that every discrimination claim, even if well-founded, is inherently more serious than every claim of a different character or where the right in question does not derive from EU law.)
45. Thirdly, Ms Monaghan pointed out that the median level of awards in the ET is quite low. She referred specifically to discrimination claims, where the figures for 2012/13 show that the percentage of awards lower than £3,000 range (as between the different protected characteristics) from just under 20% to just over 30%; and even the median awards are no higher than £7,500 (that is the figure for disability claims; race and sex are £4,831 and £5,900 respectively).8 The same observation could be made about other types of claim: the median award for unfair dismissal, for example, is £4,832, with 37% of awards under £3,000. She made the point that a claimant is less likely to risk fees of, in a type B case, £1,200 for a return that may not be much more. That point is reinforced by the fact that many losing respondents do not pay up, whether because they are insolvent or – a point emphasised by Mr Ford – because of the ineffectiveness of the available enforcement methods, so that even the fee may not be recoverable if the claim succeeds. I see this, but I do not think it can be central to the question whether a given level of fee is excessive. It was not part of Ms Monaghan's case that the fee charged should depend on the amount claimed (an idea which has some superficial attractions but is not in fact straightforward as regards claims for unliquidated compensation). As I have already said, it seems to me that the essential question is whether the claimant can, in practice, pay the fee – not whether it would be a sensible use of his or her money to do so, which would depend on many imponderables, including the likelihood of success in any given case and whether the claim is being pursued for objectives which are not purely pecuniary. If the effect of the fee regime is to make potential claimants think twice about starting proceedings for small sums, that is not axiomatically a bad thing.
(2) THE PRINCIPLES ON WHICH THE FEES ORDER CAN BE CHALLENGED
46. The present proceedings do not, unlike the Strasbourg cases discussed above, involve a claim by a particular individual or individuals that their EU or Convention rights have been breached by their being unable to afford access to the Tribunals to prosecute a particular claim. Rather, Unison's challenge is to the lawfulness of the Fees Order itself. The question arises of how the criterion of realistic unaffordability falls to be applied in that context.
47. Ms Monaghan submitted that the Fees Order would be unlawful, and would fall to be quashed, if there were "a real risk" that its provisions – that is, in effect, the level of the fees coupled with the criteria for remission – would lead to some individuals being unable to afford access to justice. She relied in particular on R (Hillingdon London Borough Council) v Lord Chancellor  EWHC 2683 (Admin),  1 FCR 1, and R (Medical Justice) v Secretary of State for the Home Department  EWCA Civ 1710; but it is necessary to refer also to R (Refugee Legal Centre) v Secretary of State for the Home Department  EWCA Civ 1481,  1 WLR 2219. I take those decisions in date order.
48. In the Refugee Legal Centre case the claimant challenged the lawfulness of the introduction of fast-track procedures for the adjudication of asylum claims. The procedures were said to be unfair because they gave applicants too little time to marshal their cases and obtain supporting evidence. One of the issues was whether it was appropriate to challenge the lawfulness of the process itself as opposed to individual instances of unfairness as and when they occurred. This Court held that it was. Sedley LJ said, at para. 7 (p. 2223D), that judicial intervention would be justified:
"... to obviate in advance a proven risk of injustice which goes beyond aberrant interviews or decisions and inheres in the system itself. In other words it will not necessarily be an answer, where a system is inherently unfair, that judicial review can be sought to correct its effects."
49. Hillingdon concerned an increase in court fees for the initiation of public law child care applications. A number of local authorities challenged the relevant order. One of the grounds was irrationality, and in that connection it was contended that the increase would or might make authorities reluctant to bring care proceedings that ought to be brought. The Divisional Court (Dyson LJ, Bennett and Pitchford JJ) acknowledged that that was an admissible ground of challenge, though it rejected it on the facts. Dyson LJ said, at para. 67 (p. 23E):
"I approach [the issue of irrationality] on the basis that the Fees Orders are unlawful if there is a real (as opposed to fanciful) risk that the increase in fees will result in local authorities not performing their statutory obligations in relation to children who are at risk."
50. In Medical Justice the challenge was to the Home Secretary's policy as regards the interval that it was appropriate to leave between the making of a removal decision under the Immigration Act 1971 and the actual removal, so as to allow the person facing removal to challenge the decision in court. Silber J, directing himself in accordance with the Refugee Legal Centre decision, had quashed the policy on the basis that "there is a very high risk if not an inevitability that the right of access to justice is being and will be infringed". An appeal to this Court was dismissed, Lord Neuberger MR holding, at para. 26, that Silber J's finding did not involve any departure from the agreed test.
51. In none of those decisions was the alleged unfairness or denial of access to justice formulated in terms of a breach of EU law or of Convention rights as such: the claimants relied straightforwardly on common law principles. But I do not see that as a matter of principle the particular source of the unlawfulness can make any difference. That is confirmed by the case-law concerning provisions of the Immigration Rules which are said to be inconsistent with article 8 of the Convention. The cases were recently exhaustively reviewed by Aikens LJ in R (MM (Lebanon)) v Secretary of State for the Home Department  EWCA Civ 985,  1 WLR 1073: see paras. 94-135 of his judgment (pp. 1105-21). At para. 133 (p. 1120F) he concluded that an immigration rule which interfered with a Convention right and which was "incapable of being proportionate and so … inherently unjustified" would necessarily be "irrational" within the meaning of the common law principles originating in Kruse v Johnson  2 QB 91.
52. The application of the principles apparent from those authorities means that the Fees Order will be unlawful if its provisions inherently create a real risk that some claimants will be denied access to justice because they cannot realistically afford the fees. There might in theory arise an interesting question whether that would be so if it were clear that the number so affected would be minimal, but that would in practice be difficult to demonstrate given that unaffordability is not a hard-edged concept; also, as will appear, the problem in this case is that there is no reliable evidence about the numbers affected in any event.
THE EVIDENCE OF UNAFFORDABILITY
53. In the Divisional Court Unison relied on two different kinds of evidence in support of its case that some claimants would be unable to afford the fees payable under the Order – (a) evidence of the financial circumstances of notional individual claimants, and (b) evidence of the fall in the numbers of claimants since the introduction of the new regime. I take the two in turn.
(a) Notional Individual Claimants
54. Unison's case in Unison 1 was at first based wholly, and always based primarily, on evidence which sought to show the relationship between the level of fees payable and the financial resources of individuals. It did not bring forward evidence of actual people: rather, it constructed examples of the financial circumstances of notional individuals who were said to be typical. Annex B to Ms Monaghan's skeleton argument for the Divisional Court set out figures illustrating the financial circumstances of, I think, eight such individuals who would, it was said, have been unable to afford to bring a type B claim in the ET under the regime introduced by the Fees Order. We have not been supplied with those figures. It became clear in the first phase of the hearing before the Divisional Court that their accuracy and/or the assumptions made in them were disputed; and at the resumed hearing in November Unison focused on only three of the eight – described as "scenarios 5, 7 and 8" – and put in evidence a witness statement from its legal officer explaining the methodology used to produce them. In short, the three scenarios are as follows:
* Scenario 5. In this scenario the notional potential claimant is a single mother with one child. She works full-time in a secretarial role in a university. She has a mortgage on her property. She has a gross monthly income of £2,271.19 (£27,255 p.a.) and "disposable monthly income", calculated according to the criteria in the original remission scheme, of £441.59. The issue fee and hearing fee for a type B claim, after remission, would represent 57% and 106% respectively of her disposable monthly income.
* Scenario 7. The notional potential claimant has a partner. Both work full-time, earning the national minimum wage. They have a child and a mortgage. Their gross monthly income is £2,497.55 (£29,970 p.a.) and their disposable monthly income is £608.43. The issue fee and hearing fee represent 41% and 82 % of that figure.
* Scenario 8. This is identical to scenario 7 except that the claimant and her partner have two children. The gross monthly income is £2,781.67 (£33,380 p.a.) and their disposable monthly income £608.43. The issue fee and hearing fee represent 45% and 93% of that figure.
55. Unison did not rely only on the high proportion of "disposable monthly income" which the fees represented. It sought also to rely on figures for the net monthly income needed to achieve "acceptable living standards" as shown in a study by the Joseph Rowntree Foundation. On the basis of the Joseph Rowntree figures none of the notional claimants would have any residual monthly income to pay the fees. What the figures were said to show was that:
"… the sample claimants chosen in scenarios 5 and 8 will not have the resources to pay the fees that they would be required to pay, even taking account of the remission scheme, and that the claimant chosen in scenario 7 is very unlikely to do so."
56. There was evidence from the Lord Chancellor disputing some aspects of those figures. I need not go into any detail. He said that the true figure for disposable income in scenario 5 was £441.59, which would yield percentages of 35% as regards the issue fee and 65% as regards the hearing fee. The Court did not feel able to resolve that issue but was prepared to accept Unison's figures for the sake of argument (see para. 33 of the judgment). As for the Joseph Rowntree Foundation figures, the Lord Chancellor in his evidence proposed a number of modifications, which would leave the claimant with a significant surplus over and above the requirements of "acceptable living standards". At para. 35 of its judgment the Court recorded:
"… The defendant rejected the applicability of [the Joseph Rowntree] study. Eventually the matter was resolved by accepting the Lord Chancellor's approach, which calculated all essential expenditure and considered the relationship between the remaining income, described as disposable monthly income, and the fees payable after remission."
It is not clear whether that means that Unison accepted the proposed modifications or that the Joseph Rowntree figures were put to one side and reliance was placed only on the calculation of disposable income according to the original remission scheme; the point was not explored before us. But it is not necessary to resolve the question. On any view Unison did not pursue its case that the claimants would have nothing left after the expenditure necessary to meet acceptable living standards.
57. The Divisional Court in its judgment reviewed the figures and concluded, at para. 40, that:
"… the combined effect of the remissions in the periods before and between the dates when fees must be paid, is that there is a sufficient opportunity even for families on very modest means, as illustrated in the three notional claimants, to accumulate funds to pay the fees. Proceedings will be expensive but not to the extent that bringing claims will be virtually impossible or excessively difficult."
It continued with the passage quoted at para. 43 above. At para. 42 it said:
"It is clear that any regime must be flexible, and have regard to the means of prospective litigants. The real difficulty lies in deciding when the level of fees imposed can properly be condemned as "excessive". The mere fact that fees impose a burden on families with limited means and that they may have to use hard-earned savings is not enough. But it is not possible to identify any test for judging when a fee regime is excessive. It will be easier to judge actual examples of those who assert they have been or will be deterred by the level of fees imposed."
58. Turning to Unison 2, Elias LJ records at para. 54 of his judgment that Unison did not in its submissions focus on the evidence about the notional individuals; rather, it relied on the evidence of the fall in the level of Tribunal claims which I consider below.
59. Before us there was likewise no focus on the cases of the three notional claimants. The grounds of appeal in Unison 1 include a general contention that the Divisional Court should have found on the evidence before it that a significant number of potential claimants would be unable to pay the fees; but the specific points pleaded are that the Court was wrong to limit itself to an approach based on "so-called 'disposable income'" (para. 2 (1)) and that it should have taken into account the evidence, preliminary though it may have been, about the fall in the number of claims since the introduction of the Fees Order (para. 2 (3)). There is exhibited to the skeleton argument a schedule – I assume originally produced for use in Unison 1 – which reviews the figures for scenarios 5, 7 and 8 (see para. 54 above) and is said to demonstrate:
"… that the cost of issuing proceedings in the ETs and EAT and the cost of pursuing such claims to a hearing will simply be prohibitive in many, many cases and the position has worsened by reason of the amendments made to the scheme by [the Remissions Order]."
But that is in the part of the skeleton which summarises the submissions below. It is not developed in the part expounding the grounds of appeal, nor did Ms Monaghan refer to the schedule in her oral submissions.
60. It is perhaps worth noting – though in truth it does not affect any issue of principle – that the class of claimant for whom the fees are said to be realistically unaffordable are not those on the lowest incomes, who will be entitled to full remission, but those whose incomes are such that they are entitled only to partial remission or are above the level at which remission ceases to be available.
(b) The Impact of the Fees Order on the Number of Claims
61. At the time of the hearing in the Divisional Court in Unison 1 only very preliminary figures were available about the impact of the Fees Order on the number of claims being brought in the ET. The Court noted that "if [the preliminary figures] are anything like accurate then the impact … has been dramatic" (para. 45). But, as already noted, its view was that any challenge on that basis was premature. At para. 46 it referred to Ms Chan's submission that "[t]he Court … should wait and see how the system will work in practice and whether indeed it will have an effect which demonstrates a breach of the [effectiveness] principle", and said that it had sympathy with that submission. It continued:
"Far better, we suggest, to wait and see whether the fears of Unison prove to be well-founded. The hotly disputed evidence as to the dramatic fall in claims may turn out to be powerful evidence to show that the principle of effectiveness, in the fundamentally important realm of discrimination, is being breached by the present regime. If so, we would expect that to be clearly revealed, and the Lord Chancellor to change the system without any need for further litigation."
62. By the time of the hearing in Unison 2, by contrast, reliable figures were available for the first year of operation of the Fees Order. These confirmed that its impact on the number of claims brought has indeed been dramatic. I am content to adopt the summary in Ms Monaghan's skeleton argument, as follows:
"Table 1.2 of the Tribunals Statistics Quarterly for October to December 2013, published on 13 March 2014, shows that, comparing the period October-December 2012 with the period October-December 2013 … :
* 79% fewer claims were accepted by the ET
* 83% fewer equal pay claims were accepted by the ET
* 77% fewer sex discrimination claims were accepted by the ET
* 84% fewer working time claims were accepted by the ET
* 80% fewer claims for a written statement of terms and conditions of employment were accepted by the ET.
The statistics released on 11 September 2014 cover the period April to June 2014. Among other things, they demonstrate that:
(1) Between April and June 2013 (before the fees were introduced), 44,334 claims were brought to the ET. Between 2014 (after the fees were introduced), just 8,540 claims were brought. That is a drop of 35,794 claims or 81%.
(2) Excluding 'multiple claims' – which is the approach the Respondent would prefer to take – between April to June 2013 (before the fees were introduced), 12,727 single claims were brought to the ET. Between April and June 2014 (after the fees were introduced), just 3,792 single claims were brought. That is a drop of 8,935 claims or 70%.
(3) Further, in this period there was an 84% drop in equal pay claims and an 81% drop in sex discrimination claims.
Overall, the falls in claims brought are as follows:
2012 v 2013
2013 v 2014
2013 v 2014
63. At para. 55 of his judgment in Unison 2 Elias LJ described those figures as "striking". He acknowledged, at para. 57, that the fall in claims might be contributed to by other recent changes in employment law, such as the introduction of a compulsory conciliation procedure; but he said that "they do not begin to explain the whole of this very dramatic change". He continued:
"58. Ms Chan's basic submission, however, is that whatever the statistics say they cannot of themselves demonstrate that the principle of effectiveness has been infringed. It is not legitimate to infer that some litigants cannot pay from the fact that a significant number do not pay. Ms Chan accepts that the imposition of a fee will necessarily deter some litigants from taking their cases but contends that there are likely to be a variety of reasons for this. Some workers who in the past may have pursued a weak case, if only in the hope of securing a small settlement in their favour, will now be reluctant to do so because of the risk of having to pay fees if the case goes to the tribunal. Others will quite properly choose to spend their limited resources in other ways rather than gamble on litigation. Ms Chan also points out that the most seriously disadvantaged are covered by the remission scheme; and that in general a successful litigant can recover the fees against the unsuccessful party. So a party with a good case can pursue it with some confidence.
59. The claimant and the intervener challenge this analysis. They do not accept that the payment of fees is likely to encourage settlement. Indeed, they submit that this is less likely if the employer considers that the employee will not choose to litigate because of the fee burden: there will then be no incentive for the employer to compromise. They submit that even if one allows for the fact that there may have been some reduction because of the introduction of compulsory conciliation – which they submit on any view would be very small - that does not begin to explain the very significant reduction. It must be the case that some litigants who actively wish to take proceedings simply cannot afford to do so because they are not entitled to remission of the fee and earn too little; no other sensible inference is possible on these figures. The income level at which relief from paying fees is lost is set too low; the effect is that for many low earners the right of access to the tribunal is indeed illusory.
60. I see the force of this submission and I suspect that there may well be cases where genuinely pressing claims on a worker's income will leave too little available to fund litigation. But the difficulty with the way the argument has been advanced is that the court has no evidence at all that any individual has even asserted that he or she has been unable to bring a claim because of cost. The figures demonstrate incontrovertibly that the fees have had a marked effect on the willingness of workers to bring a claim but they do not prove that any of them are unable, as opposed to unwilling, to do so.
I have quoted the next paragraph at para. 43 above. Elias LJ then continued:
62. In my view, the court can only properly test the argument if there are actual cases which will enable the court to review the income and expenditure of a particular individual or individuals and apply the effectiveness principle in that concrete situation, as Moses LJ emphasised in the earlier proceedings."
64. Foskett J agreed with Elias LJ, but he added a brief judgment of his own, which included the following:
"96. As Elias LJ has recorded (at paragraphs 55-56), the effect of the introduction of the new regime has been dramatic. Indeed it has been so dramatic that the intuitive response is that many workers with legitimate matters to raise before an Employment Tribunal must now be deterred from doing so because of the fees that will be demanded of them before any such claim can be advanced. For my part, I would anticipate that if the statistics upon which reliance is placed in support of this application were drilled down to some individual cases, situations would be revealed that showed an inability on the part of some people to proceed before an Employment Tribunal through lack of funds which would not have been the case before the new regime was set in place. However, that assessment has to be seen as speculative until convincing evidence to that effect is uncovered. If it is, of course, the Lord Chancellor would doubtless feel obliged to address it.
97. Elias LJ has referred to the way in which the statistics have been deployed in this application and has identified the fact that no evidence from any individual who has been affected adversely by the new regime (in the sense that it is now virtually impossible or extremely difficult to proceed through lack of funds) has been given: see paragraph 61 above. Whilst the analogy is not exact, it seems to me to be akin to trying to prove the causation of damage in an individual case by reference to statistical evidence: see the discussion at paragraph 2-28 in Clerk & Lindsell on Torts, 21st ed. There can be little doubt that the statistics relied upon in this case raise a legitimate question about the operation of the new regime, but they do not provide the answer to that question.
98. As it seems to me, before the court could begin to act it would need to be satisfied that a more than minimal number of people with arguably legitimate claims would find it virtually impossible or excessively difficult to bring such matters before an Employment Tribunal because of the fees that would require to be paid."
65. Before us, Ms Monaghan submitted that the decline in the number of ET claims, on its own and without more, gives rise to an "irresistible" inference that the new fee regime is making it in practice impossible, or excessively difficult, for a significant number of claimants to enforce their rights.
66. As noted above, there is an application before us for the figures summarised at para. 62 above to be admitted as evidence in the Unison 1 appeal. The question would seem to be academic since they are before us in any event in Unison 2. But for what it is worth, I do not think that they are admissible in the Unison 1 appeal. The challenge is to the Lord Chancellor's decision to make the Fees Order in July 2013. I do not see how that decision can be impugned on the basis of information that not only was not but could not have been available to him at that date. It is another matter as regards Unison 2, since Ms Monaghan can and does argue that it is unlawful for the Lord Chancellor to maintain the Order in effect once it has become clear that it is denying many claimants access to justice.
DISCUSSION AND CONCLUSION
67. It is quite clear from the comparison between the number of claims brought in the ET before and after 29 July 2013 that the introduction of fees has had the effect of deterring a very large number of potential claimants. However, that by itself does not evidence or constitute a breach of the effectiveness principle. It is inevitable that potential claimants will be more willing to embark on litigation when it is free than when they have to make a payment upfront, which there is no certainty of recovering; and the introduction of fees was accordingly bound to have such an effect. But it is well-established that the charging of court fees is not in itself objectionable. To the extent that the decline in claims is due to potential claimants who would previously have chosen to litigate making a choice not to do so, that does not assist Unison. Applying the principles discussed at paras. 36-46 above, the question is whether the introduction of the fees regime has in at least some cases made it not simply unattractive but in practice impossible to pursue a claim. This is the point made by Elias LJ at para. 61 of his judgment, quoted at para. 43 above, and I agree with it.
68. I have found this part of the case troubling. Like both Divisional Courts, I have a strong suspicion that so large a decline is unlikely to be accounted for entirely by cases of "won't pay" and that it must also reflect at least some cases of "can't pay"; and I have accordingly been tempted by Ms Monaghan's submission that the figures speak for themselves. But in the end I do not think that that is legitimate. The truth is that, looked at coolly, there is simply no safe basis for an untutored intuition about claimant behaviour or therefore for an inference that the decline cannot consist entirely of cases where potential claimants could realistically have afforded to bring proceedings but have made a choice not to. But in fact the difficulty goes further than that. Even if it really were an irresistible inference from the decline in claims that at least some potential claimants could not realistically afford the fees, there is still no basis for forming any reliable view about the numbers of such cases, or how typical they may be; and, for reasons which will appear, that is an important matter. In my view the case based on the overall decline in claims cannot succeed by itself. It needs to be accompanied by evidence of the actual affordability of the fees in the financial circumstances of (typical) individuals. Only evidence of this character will enable the Court to reach a reliable conclusion that that the fees payable under the Order will indeed be realistically unaffordable in some cases.
69. It was the view of the Court in Unison 2 that that evidence could not be provided by reference only to the cases of notional individuals, and that Unison should have adduced evidence of actual potential claimants who are said to have been unable to afford to pay the fees: see paras. 60 and 62 of Elias LJ's judgment and paras. 96-98 of Foskett J's (paras. 63-64 above). The Court in Unison 1 made similar observations (see para. 42, quoted at para. 57 above), though in more tentative terms. I disagree. I see no reason in principle why well-constructed cases of notional individuals could not be used to assist in proving that the fees would be realistically unaffordable for at least some typical claimants.
70. That presents a forensic difficulty for Unison. Evidence of that character was of course put forward in Unison 1. But, as set out at para. 59 above, the reliance on it in the grounds of appeal and skeleton argument was limited, and Ms Monaghan did not advance any oral submissions in relation to it: as in Unison 2, she focused on the overall decline in claims, no doubt judging that that was sufficient to prove her case. No attempt was made to argue that the Divisional Court's decision about what conclusions could be drawn from the figures for the three notional individuals was wrong. I am reluctant to base any conclusions on the evidence about those individuals in circumstances where the question of what it established not the subject of argument before us.
71. Notwithstanding that, I have considered the figures in the schedule to Ms Monaghan's skeleton argument. This embodies two exercises. First, it expresses the fees payable under the new remission scheme as a percentage of disposable income (as calculated for the purpose of the old remission scheme – so this is something of a hybrid): I have set out the percentages at para. 54 above. Secondly, in scenarios 5 and 8 it sets out the amount of income left after "acceptable living standards" have been met according to the Joseph Rowntree criteria: adopting the modified version of those criteria proposed by the Lord Chancellor, that shows that the claimant in scenario 5 had about £200 left per month out of which to pay the fees and the claimant in scenario 8 about £590. The Divisional Court did not regard those figures – whether including the "modified" Joseph Rowntree figures or not – as showing that the fees in question were realistically unaffordable for those claimants, having regard to their opportunity to accumulate the necessary sums over a period of months before the issue of the claim and before the hearing: see para. 40 of its judgment, quoted at para. 57 above. I am not prepared definitively to endorse that conclusion in circumstances where we heard no developed challenge to it, but it is not obviously wrong.
72. There is, however, a further point which would be decisive against the appeal on this issue, even if it were sufficiently proved that at least some individuals cannot realistically afford to pay the fees on the basis of the current remission criteria. This is not a claim arising out of the circumstances of a particular case but a challenge to the lawfulness of the Fees Order as a whole, which falls to be judged on the basis discussed at paras. 46-52 above. In this context it is necessary to consider the effect of paragraph 16 of Schedule 3: see paras. 20-24 above. If claimants of the kind postulated – that is, those who are realistically unable to afford to pay the fees but who fall outside the primary criteria for remission – can obtain remission by establishing "exceptional circumstances" under that provision, then no breach of the effectiveness principle would arise, because the regime would not inherently result in them being unable to bring Tribunal proceedings. Elias LJ made essentially the same point at para. 63 of his judgment in Unison 2, which reads as follows:
"A related problem is that the relief sought is to quash the relevant regulations. But in my view the Lord Chancellor would be entitled to know in what circumstances the scheme is considered to be defective in order to remedy it. I appreciate of course that it is not for the court to draft a lawful scheme; that is for the Lord Chancellor. But here the court would be saying no more than that the inevitable inference from the statistics is that the scheme leaves an indefinable and undefined number of people with no effective way to redress wrongs. The difficulty is compounded by the fact that the Lord Chancellor has a discretion to relieve a claimant from the obligation to pay fees in exceptional circumstances. It is true that we have heard little about this discretion or when it may be exercised. But if the evidence were to suggest that the number of claimants for whom the right of access is illusory is likely to be very small, invoking the discretion might be a way of dealing with the problem whilst leaving the scheme itself intact. Quashing the scheme would then be inappropriate. … ."
73. I accept that that argument cannot be taken too far. Paragraph 16 would not save the lawfulness of the Fees Order if it had to be pressed into service to afford remission to claimants where the circumstances could not on any view be described as exceptional: it would be illegitimate to construe it so that the exception overwhelmed the rule. Having said that, in my view the Lord Chancellor in operating paragraph 16, and the Court in reviewing its scope in the present context, should take a reasonably broad view of what might constitute "exceptional circumstances": it need not be confined to situations which are wholly idiosyncratic or very rare. The Order must so far as possible be interpreted conformably with the effectiveness principle – that is, in such a way as to ensure that all claimants who cannot realistically afford to pay the full fees are entitled to remission. The Court should not take the drastic step of striking down the Fees Order if the EU and Convention rights of claimants can on a conforming interpretation be accommodated within its own terms.9 In my view that would be so even if the numbers achieving remission by this route turned out to be rather larger than Elias LJ's reference to "very small" might suggest. Setting the criteria for remission inevitably involves a large element of judgement, with room for reasonable differences of view as to their precise calibration; and I see no reason why paragraph 16 should not in principle cover cases which arguably, with hindsight, ought to have been accommodated by setting the criteria rather less restrictively. It is futile to try to define in the abstract at what point the number of claimants having to rely on paragraph 16 is so large that the circumstances can no longer fairly be described as exceptional. The point is that the evidence adduced affords no way of assessing the numbers of those who are unable realistically to afford the fees on the basis of the primary remission criteria.
74. Ms Monaghan did not advance any challenge to the terms of the Lord Chancellor's guidance about the criteria to be applied in determining an application under paragraph 16 (see paras. 22 and 23 above). In those circumstances I will not comment on them, save to say that neither the reference to "unexpected events" nor the particular examples given should be treated as definitive, and that those considering such applications should appreciate that the effectiveness principle requires that remission should be afforded to claimants who cannot realistically afford to pay the fee. Any particular decision to refuse remission under paragraph 16, maintained on appeal, in a case where the claimant was not in fact able to afford the fee could in principle be challenged by judicial review. Ms Monaghan observed, and I would accept, that that would rarely be a realistic remedy; but I do not believe that the risk of an aberrant decision which it would not then be easy to challenge is by itself a sufficient ground for striking down the Fees Order.
CONCLUSION ON BREACH OF THE EFFECTIVENESS PRINCIPLE
75. I would accordingly hold that it has not been proved that the Fees Order breaches the principle of effectiveness. I would, however, say this. On 11 June 2015 the Lord Chancellor announced a post-implementation review, which would "consider how effective the introduction of fees has been in meeting the original financial and behavioural objectives while maintaining access to justice [my emphasis]": it had in fact been made clear before the Divisional Courts that such a review would be conducted in due course. The fact that the evidence put before this Court has not satisfied me that there has been a breach of the effectiveness principle should not, and I am sure will not, preclude the Lord Chancellor from making his own assessment, on the basis of the evidence to which he will have access, on that question. The decline in the number of claims in the Tribunals following the introduction of the Fees Order is sufficiently startling to merit a very full and careful analysis of its causes; and if there are good grounds for concluding that part of it is accounted for by claimants being realistically unable to afford to bring proceedings the level of fees and/or the remission criteria will need to be revisited.
(B) INDIRECT DISCRIMINATION
THE RELEVANT PRINCIPLES
76. It is Unison's pleaded case that the Fees Order is unlawfully discriminatory by reference to EU law, article 14 of the Convention and section 29 (6) of the Equality Act 2010. But Ms Monaghan focused her submissions on EU law, and she did not suggest that anything relevant could be gained by relying in addition or alternatively on the Convention or the Act. She said that the relevant principles appeared from two decisions of the ECJ concerned with indirectly discriminatory legislation – O'Flynn v Adjudication Officer (C-237/94),  ECR I-2617, and Borawitz v Landesversicherungsanstalt Westfalen (C-124/99),  ECR I-7293. I should start by setting out the passages from those two authorities on which she relied.
77. O'Flynn was concerned with a provision of UK social security legislation under which a claimant was entitled to assistance with funeral costs only if the funeral took place in the UK. The claim was brought by an Irish migrant worker resident in the UK whose son had died here but been buried in Ireland and whose claim for a funeral payment had accordingly been refused. The ECJ held that the condition was indirectly discriminatory against non-UK nationals. It stated the applicable principles at paras. 17-18 of its judgment as follows:
"17. The Court has consistently held that the equal treatment rule laid down in Article 48 of the Treaty and in Article 7 of Regulation No 1612/68 prohibits not only overt discrimination by reason of nationality but also all covert forms of discrimination which, by the application of other distinguishing criteria, lead in fact to the same result (see inter alia Case 152/73 Sotgiu v Deutsche Bundespost  ECR 153, paragraph 11; Case C-27/91 URSSAF v Le Manoir  ECR I-5531, paragraph 10; Case C-111/91 Commission v Luxembourg  ECR I-817, paragraph 9; and Case C-419/92 Scholz v Opera Universitaria di Cagliari  ECR I-505, paragraph 7).
18. Accordingly, conditions imposed by national law must be regarded as indirectly discriminatory where, although applicable irrespective of nationality, they affect essentially migrant workers (see Case 41/84 Pinna v Caisse d'Allocations Familiales de la Savoie  ECR 1, paragraph 24; Case 33/88 Allué and Another v Università degli Studi di Venezia  ECR 1591, paragraph 12; and Le Manoir, paragraph 11) or the great majority of those affected are migrant workers (see Case C-279/89 Commission v United Kingdom  ECR I-5785, paragraph 42, and Case C-272/92 Spotti v Freistaat Bayern  ECR I-5185, paragraph 18), where they are indistinctly applicable but can more easily be satisfied by national workers than by migrant workers (see Commission v Luxembourg, paragraph 10, and Case C-349/87 Paraschi v Landesversicherungsanstalt Württemberg  ECR I-4501, paragraph 23) or where there is a risk that they may operate to the particular detriment of migrant workers (see Case C-175/88 Biehl v Administration des Contributions  ECR I-1779, paragraph 14, and Case C-204/90 Bachmann v Belgium  ECR I-249, paragraph 9)."
Those paragraphs reflect the fuller reasoning of A-G Lenz, at paras. 14-29 of his opinion, from which it is clear that the Court intended to reject a contention advanced by the UK that the provisions in question would only be discriminatory if "a substantially higher number of nationals of other member states than of nationals of the member state [were] affected" (see para. 18): there was no reason to suppose that there were not as many UK nationals as non-UK nationals who were buried abroad. A-G Lenz acknowledged that such a numerical approach might indeed be necessary in some kinds of case, but he said that that was not the case where the nature of the measure was such that it was inherently more likely to affect persons in the protected category. That was so in the instant case because it was plainly inherently more likely that nationals of member states would be buried outside the UK than UK nationals: see in particular para. 27 of his opinion.
78. Borawitz was concerned with a German regulation which applied differential rules about pension payments to workers resident in Germany and those resident in another member state. This was held to be unlawfully indirectly discriminatory against non-German nationals. At para. 25 of its judgment the ECJ recapitulated what it had said in O'Flynn. After referring at para. 26 to the need for objective justification, it continued, at para. 27:
"It is clear from that body of case-law that, unless it is objectively justified and proportionate to its aim, a provision of national law must be regarded as indirectly discriminatory if it is intrinsically liable to affect the nationals of other Member States more than the nationals of the State whose legislation is in point and if there is a consequent risk that it will place the former at a particular disadvantage (see, to that effect, Case C-57/96 Meints v Minister van Landbouw, Natuurbeheer en Visserij  ECR I-6689, paragraph 45)."
79. I am bound to say that I do not regard those authorities as the best starting-point. Para. 18 of the judgment in O'Flynn illustrates a series of particular ways in which legislation may be indirectly discriminatory but it does not state the principle underlying them. There is some such statement in para. 17 but it is very broadly expressed and uses obsolete terminology: the language of "overt" and "covert" discrimination has long been regarded as inapt10. The opinion of A-G Lenz is more helpful but it also uses the same out-of-date terminology.
80. However, the underlying principle is sufficiently well-established. The Directives identified at para. 44 above each define indirect discrimination in substantially the same terms. Taking 2006/54/EC by way of example, article 2.1 defines indirect discrimination as
"… where an apparently neutral provision, criterion or practice [in the jargon, a "PCP"] would put persons of one sex at a particular disadvantage compared with persons of the other sex, unless that provision, criterion or practice is objectively justified by a legitimate aim, and the means of achieving that aim are appropriate and necessary."
In the case of legislation which is said to be indirectly discriminatory, the impugned provisions will constitute the PCP, and the issue will be whether they put persons in one of the groups protected by EU law at "a particular disadvantage" compared with persons not in that group. That approach is in fact clearly reflected, albeit not spelt out, in the judgments in O'Flynn and Borawitz: on the facts of those cases, and in the other examples given in O'Flynn, the terms of the impugned legislation necessarily put non-nationals, or their families, at a "particular" disadvantage compared with nationals of the legislating state. Indeed the terminology of particular disadvantage is used in para. 27 of the judgment in Borawitz (and there is reference in O'Flynn to "particular detriment"). The essential question in this case is thus whether the effect of the Fees Order is to put persons in a protected category at a particular disadvantage compared with persons not in that category; and if so whether that disadvantage can be objectively justified.
81. Ms Monaghan was at pains to emphasise that EU law as stated in O'Flynn and Borawitz was "flexible" in its approach to indirect discrimination, and specifically that it is not necessary in every case that "particular disadvantage" should be established by statistical analysis. That is no doubt so, as appears particularly from the opinion of the Advocate-General. The current definition of indirect discrimination is indeed more flexible than formulations found in earlier EU legislation and case-law, and also in the original UK anti-discrimination statutes (though I am not sure that the older approach was as rigid as is sometimes suggested). That was authoritatively recognised in Chief Constable of West Yorkshire Police v Homer  UKSC 15,  ICR 704 – see per Lady Hale at para. 14 (p. 709 G-H). What kind of analysis is necessary in order to demonstrate "particular disadvantage" will depend on the circumstances of the particular case. But the change has not affected the fundamental underlying concepts. As Ms Monaghan accepted, the concept of a "particular" disadvantage necessarily entails a comparative exercise, whereby the disadvantage suffered by the members of the protected group is compared with the disadvantage suffered by comparable people outside the group to see if there is a disparate impact. Recognising the need for flexibility in the way in which indirect discrimination may be proved does not remove the need for a clear analysis of the comparative disadvantage being alleged in any particular case.
THE WAY THE CASE IS PUT
82. It is Unison's case that the Fees Order is discriminatory against persons in each of the protected categories recognised in EU (and domestic) law. But Ms Monaghan said that for simplicity of presentation she would focus on women as the paradigm protected category, as she had also done before the Divisional Court in Unison 2. Mr Ford took the same approach. The implication is that the analysis is interchangeable whether one considers the effect on persons with the protected characteristics as a whole or focuses on any particular characteristic. I am not sure that that is entirely right, but I need not explore the point further.
83. Despite the usual persuasiveness of Ms Monaghan's advocacy, there was some lack of precision in identifying how her case on indirect discrimination was put. (One consequence of this was an uncertainty, which was not definitively resolved in her oral submissions, as to whether it was her case that the fees regime as a whole was unlawfully discriminatory or only the charging of a higher level of fee for type B cases.) In my view the best starting-point is Elias LJ's judgment in Unison 2, where he analysed what he understood her case to be into three distinct variants. Ms Monaghan recorded that analysis without disagreement in her skeleton argument.
84. I should say something about how the indirect discrimination case went in Unison 1. As appears from paras. 76-78 of the judgment, it was bedevilled by confusions about the statistical evidence. There was no evidence about the breakdown of claimants by gender or any other protected characteristic for type A and type B claims under the new regime, and the parties had to proceed on the basis of statistics of some age, produced for other purposes, which by definition did not reflect the post-implementation position. Those difficulties were compounded by the late production of further evidence. The Court was prepared, with some hesitation, to accept that some disparate impact on women was probably established (para. 84), and it briefly reviewed some of the questions relevant to justification (paras. 85-87); but in the end it decided that the material before it did not permit it to come to any fair conclusion. Although Unison is critical of it for taking that course, the result is that the only definitive consideration is that of Elias LJ in Unison 2, and the focus of the appeal has inevitably been on his reasoning and conclusions.
85. Elias LJ accepted, at para. 68 of his judgment, that the available statistics appeared to show that the proportion of type B claimants who were women was higher than the proportion of type A claimants. It appears that he had in mind figures in Ms Monaghan's skeleton argument asserting that 54% of type B claimants were women but only 37% of type A claimants. Mr Barr expressed some "puzzlement" about Elias LJ's conclusion, because the type B figure is different from that which he appeared to accept in relation to variant (3): see para. 104 below. But this was not the subject of a Respondent's Notice and in any event even the revised percentage (45%) would leave a disparity. I think we should proceed on the basis that there was indeed a disproportion of the kind which Elias LJ accepted.
86. Elias LJ accepted that that was a situation which required to be justified and Mr Barr did not argue otherwise: that being so, I need not spell out the details of the analysis, save to say that the PCP complained of is the requirement to pay the specified level of fee in order to bring a type B claim. He pointed out, however, that:
"the logic of this argument is not that fees cannot be charged or that the scheme should be quashed, which is the relief sought; rather it is that women being indirectly discriminated against for level B claims should not have to pay more than level A fees"
(para. 68). As regards this variant at least, Ms Monaghan accepted that that is plainly correct.
87. Elias LJ held that the disproportion in question had been justified. His reasoning is at para. 69, as follows:
"The issue here is whether the difference in the fee is justified rather than whether any fee is justified. The rationale for the distinction between category A and B cases is that those subject to level A fees are in general likely to take less time than claims falling within category B and therefore use fewer resources. Ms Monaghan submitted that there is no direct evidence of this and that the court should not simply accept counsel's assertion to that effect. I do not accept that. In a document produced by HM Courts and Tribunals Service giving information about the fees it is expressly stated that "Type A claims tend to be more straightforward for the Tribunal to deal with, and so have lower fees." Moreover, there is clearly some rationale for the different funding arrangements for groups A and B, and in my view the explanation given is consistent both with the reason for imposing the fees in the first place and with the nature of the claims falling within the two groups. In my judgment, it is legitimate to fix the fees by reference to the service - in the sense of court resources - provided. It is true that the scheme adopts bright line rules; some level A claims will take longer than some level B claims and vice versa. But it is legitimate in circumstances like this to regulate by reference to the cost of the service in standard cases. I would therefore reject this ground."
88. I should say that there is a further section of Elias LJ's judgment, at paras. 82-91, which is headed "Justification" and which appears to be intended to address the question of justification rather more generally in respect of any disparate impact which the Fees Order might have been shown to have on members of protected classes. That is apparent from its conclusion, at para. 90, which reads:
"I have no doubt that each of the objectives relied upon in this case is a legitimate one and that the scheme taken overall, particularly having regard to the arrangements designed to relieve the poorest from the obligation to pay, is justified and proportionate to any discriminatory effect."
It is not therefore primarily directed at the justification for the specific disparate impact which is the subject of variant (1); but there is some overlap, as will appear below.
89. Ms Monaghan challenged Elias LJ's reasoning on variant (1) on essentially three bases, which I take in turn.
90. First, she submitted that his argument depended on accepting that the different levels of cost to HMCTS associated with type A and type B claims were a proper justification for the difference in the fees charged, whereas it is established that "cost" alone is not an admissible justification in cases of discrimination. Mr Ford made the same point on behalf of the Commission. In support of that submission she relied on the judgment of the Supreme Court, delivered by Lord Hope and Lady Hale, in Ministry of Justice v O'Brien  UKSC 6,  ICR 499. The case concerned the denial of pension rights to judges who sit part-time: it was considered by the Supreme Court following a reference to the ECJ (C 393/10),  ICR 955. At para. 63, under the heading "Cost", the Court says:
"The Ministry accept that cost alone cannot justify discriminating against part-time workers. But they argue that "cost plus" other factors may do so. This is a subtle point which is not without difficulty."
At paras. 64-68 it reviews the ECJ case-law, starting with MA (Roks) v Bestuur Van de Bedrijfsvereniging voor de Gezondheid (C-343/92),  2 CMLR 325. At paras. 69-70 it says:
"69. Hence the European cases clearly establish that a Member State may decide for itself how much it will spend upon its benefits system, or presumably upon its justice system, or indeed upon any other area of social policy. But within that system, the choices it makes must be consistent with the principles of equal treatment and non-discrimination. A discriminatory rule or practice can only be justified by reference to a legitimate aim other than the simple saving of cost. No doubt it was because the CJEU foresaw that the Ministry would seek to rely upon considerations of cost when the case returned to the national courts that it took care to reiterate that "budgetary considerations cannot justify discrimination" (para 66).
70. Our attention was drawn to some domestic authorities, and in particular Woodcock v Cumbria Primary Care Trust  EWCA Civ 330,  ICR 1126. This was an age discrimination case, in which the claimant complained that the trust had deliberately failed to comply with a requirement to consult before declaring him to be redundant, so that his employment would cease before he reached the age which would trigger a higher severance payment. The Court of Appeal held that the dismissal notice was not served with the simple aim of dismissing him before his 49th birthday but in order to give effect to a genuine decision that his position was redundant. It was justifiable to implement that decision in a way which saved money. This court must, however, take its guidance from the jurisprudence of the CJEU, and in particular the guidance which we have been given in this very case. In the circumstances it is unnecessary for us to express a view upon whether the case of Woodcock was rightly decided."
91. There is in discrimination litigation a good deal of bandying about of the language of "cost", which is said never to be an admissible justification, and "cost plus", in which the presence of some other factor appears magically to legitimate partial reliance on cost considerations. I have expressed elsewhere11 my concerns about this crude dichotomy; and I respectfully agree with the Supreme Court in O'Brien that the issue is one of some subtlety. But we were not addressed about the underlying principles, and I do not think it would be satisfactory to try to grapple with them here. Fortunately, I do not believe that it is necessary. I do not believe that the present case falls into the kind of category addressed in the ECJ authorities summarised in O'Brien or indeed in O'Brien itself. It is necessary to proceed, as regards this variant, on the basis that the charging of fees is not in itself discriminatory against women (or any other protected class). We are not therefore concerned with the justification for charging fees at all, which might be characterised as primarily "budgetary". Rather, the challenge is a more limited one, namely to the charging of a higher level of fees for what are perceived to be more complex cases. The underlying rationale is that there ought to be a relationship between the level of the fee and the degree of the demand on the Tribunals' resources. I do not regard that as invoking "cost" or "budgetary considerations" in the same sense as the Ministry of Justice was said to be doing when it argued that it was unduly expensive to accord pension rights to part-time judges. Given that it is legitimate to charge tribunal fees in the first place, a graduated level of fees which reflects the extent to which the resources of the tribunal are engaged is no more than an application of ordinary principles of economic efficiency, which has regularly been accepted as relevant to justification: it is not necessary to cite more than the celebrated decision of the ECJ in Bilka-Kaufhaus GmbH v Weber von Hartz (C-170/84),  ICR 110 – see para. 36 (p. 126 E-G). To avoid any possible misunderstanding, I am not to be taken as saying that economic efficiency will always be a sufficient justification for a PCP having a discriminatory impact, but Ms Monaghan's challenge was to the admissibility of this consideration at all.
92. My conclusion on this issue is in line with that of Elias LJ when he addressed it in the context of his broader discussion of justification to which I have referred above. Having held, at para. 89, that any cost-based justification was permissible on the "cost plus" analysis, he continued:
"But I would not in fact describe the first objective as costs saving. This is not a case of government refusing to correct discrimination because it would be too expensive. Rather it is more accurately characterised as requiring a contribution towards the cost of running the Tribunal Service, charging equal amounts from all who bring claims within class B."
93. Ms Monaghan's second challenge was that it was illegitimate to seek to justify the differential charge by reference to a "bright line" rule, as Elias LJ had done. She relied on the decision of the House of Lords in AL (Serbia) v Secretary of State for the Home Department  UKHL 42,  1 WLR 1434. In that case the Home Office had adopted a policy under which some categories of asylum-seekers with families were granted leave to remain outside the Immigration Rules. The policy was challenged on the basis that it involved a breach of article 14 of the Convention because it was discriminatory against young single adults. The discrimination in question was direct discrimination, which can in the Convention jurisprudence in principle be justified. Ms Monaghan relied on observations by Lady Hale at paras. 35 and 45 of her judgment to the effect that discrimination on what the ECtHR regarded as "suspect" grounds – which would embrace the protected characteristics defined in EU law – are more difficult to justify than discrimination on other grounds (such as were involved in that case). That is no doubt true, but I do not see what it has to do with the present case, which is concerned with indirect discrimination. The only mention of "bright lines" is at para. 44 (p. 1449H), where Lady Hale notes that the Secretary of State had chosen not to adopt an approach based on the date on which applicants had entered the country and had observed that:
"It is accepted that bright lines of this sort, even if they produce what appear to be arbitrary distinctions between one case and another, are often necessary and can be justified: see, e g, the age rules in (R (Reynolds) v Secretary of State for Work and Pensions sub nom R (Carson) v Secretary of State for Work and Pensions  1 AC 173)."
Ms Monaghan sought to conflate those two passages, arguing that a "bright line" rule cannot be justified where the discrimination in question relates to a protected characteristic. Whether the passages are read separately or together, they do not support such a conclusion. The fact that the distinction between type A and type B claims only reflects their different levels of demand on the ET's resources in a rough-and-ready way is no doubt a relevant consideration in assessing whether those greater demands justify the disparate impact on female claimants; but it is no more than one consideration.
94. Ms Monaghan's third challenge picked up a point made in particular by Mr Ford in his written submissions on behalf of the Commission. He referred to the decision of the CJEU in Steinicke v Bundesanstalt für Arbeit (C-77/02),  1 CMLR 49. That case concerned alleged indirect discrimination against part-time workers, which the employer sought to justify on the basis that the measures in question encouraged recruitment. The Court observed at para. 64 of its judgment (p. 1371) that:
"Mere generalisations concerning the capacity of a specific measure to encourage recruitment are not enough to show that the aim of the disputed provisions is unrelated to any discrimination on grounds of sex or to provide evidence on the basis of which it could reasonably be considered that the means chosen are or could be suitable for achieving that aim … ."
She submitted that the justification based on the greater demand placed on the ET's resources by type B claims amounted to a "mere generalisation".
95. I do not believe that that is a fair criticism. The extent to which it is legitimate to rely on a general proposition in support of a claimed justification will vary according to the subject-matter. Ms Monaghan's submission exemplifies the too common tendency of advocates to pick up an observation made (particularly by the CJEU) in the circumstances of a particular case and treat it as a lapidary pronouncement to be unthinkingly applied in all other cases. The fact that the proposition relied on by the employer in Steinicke may have been inadequate unless supported by specific evidence does not mean that the same is true of the proposition accepted by Elias LJ in the present case. The statement from HMCTS which he accepted to the effect that type A claims are typically more straightforward to deal with is in no way surprising or implausible: on the contrary, it is entirely in accordance with common experience, as Elias LJ was peculiarly well placed to recognise. If Unison had advanced a positive case that the statement was wrong, some more specific evidence to support it might have been needed. But it did not.
96. Accordingly I do not believe that any flaw has been shown in Elias LJ's reasoning or conclusion on this variant of Unison's indirect discrimination challenge.
97. Elias LJ described this variant as "that there is discrimination against those who are bringing discrimination claims". It seems that he understood the challenge to be based on the fact that "the proportion of women who bring discrimination claims is greater than the proportion of men" (loc. cit.) – or, to put the same thing another way, the majority of discrimination claims are brought by women. The statistics apparently showed that 58% of all discrimination claims are brought by women, no doubt because a high proportion of such claims are claims of sex discrimination, which are almost always brought by women.
98. Elias LJ's response to that way of putting the case, at para. 71 of his judgment, was as follows:
"… I do not think that to select a sub-group of cases within category B is a legitimate way to seek to establish indirect discrimination. It is necessary to test any potentially adverse effect of the provision, criterion or practice (PCP) by focusing on all those who are subject to it, the overall pool to whom the PCP is applied. It is not legitimate to take a self-selected group. That simply distorts the true effect of the PCP. Moreover, it yields bizarre results. If there is an adverse impact on women for discrimination claims, there must be a corresponding adverse impact on men for all non-discrimination claims (and apparently there is for unfair dismissal cases, for example). Ms Monaghan's riposte is to say that there may well be indirect discrimination against men in those cases, and that this would need to be justified too. But on that analysis, even if the PCP operated to advantage one sex overall, by a judicious selection of a particular subgroup where the claimants were predominantly of the other sex, it could be shown that the rule indirectly disadvantaged the group predominantly advantaged by the PCP as a whole. By choosing a subgroup which is in practice predominantly of one sex - say nurses or building workers - or by selecting claims typically made by one sex rather than the other, as has been done here, it would be possible to show that there was in fact indirect discrimination being practised in a whole variety of ways and each distinct type would have to be justified. I do not accept that the concept of indirect discrimination has such unacceptable and arbitrary consequences."
In the following paragraphs he supported that analysis by reference to the decisions of this Court in University of Manchester v Jones  ICR 474 and London Underground v Edwards (no. 1)  ICR 574, which were concerned with the definition of the relevant "pools" for the purpose of the Sex Discrimination Act 1975. He also referred to Lady Hale's observation in Rutherford v Secretary of State for Trade and Industry (no. 2),  UKHL 19,  ICR 785, that "indirect discrimination cannot be shown by bringing into the equation people who have no interest in the advantage or disadvantage in question" (see para. 82 of her judgment, at p. 808A).
99. I should summarise in my own words what I understand to be the gist of that reasoning. The PCP relied on in variant (2) is, again, the provision of the Fees Order that claimants bringing type B claims must pay the higher level of fee. That "disadvantages" all type B claimants, but it is said to particularly disadvantage women because more women than men bring discrimination claims. Elias LJ's point is that that is not the relevant disproportion: the only relevant disproportion would be between women and men bringing type B claims, since that is the group suffering the disadvantage. Type B claims are not of course limited to discrimination claims: most significantly, they also include claims of unfair dismissal. No gender disproportion is alleged as regards the group as a whole, and accordingly the higher level of fee cannot be said to particularly disadvantage women.
100. In my view that is right. Ms Monaghan contended that Elias LJ's approach was over-analytical, and that it was no longer appropriate to refer, as he had done, to authorities concerned with the definitions of indirect discrimination in the earlier UK legislation in the light of the more flexible approach which EU law permits; and Mr Ford made the same point for the Commission. But, as I have already made clear, the different formulation now in use has not entailed any change in the essential ingredients of an indirect discrimination claim, and any greater flexibility which it may permit affects only the means of proving those ingredients. Pre-Directive domestic case-law may need to be handled with care to the extent that it is concerned with the particular language of the older legislation, but it may still contain useful discussions or illustrations of basic principle; and there was nothing wrong in Elias LJ referring to it.
101. Mr Ford in his written submissions referred to the decision of the CJEU in HK Danmark v Dansk Almennytiggt Boligselskab (C-335/11),  ICR 851, in which a provision of Danish law which allowed employers to dismiss an employee after a specified period of sickness absence was held to be indirectly discriminatory against disabled employees on the basis that it is inherently more likely that such employees would be absent sick for long periods than non-disabled employees. He submitted that the present case was analogous. I do not agree. Women are no doubt inherently more likely than men to bring sex discrimination claims; but the requirement to pay type B fees is not applied only to sex discrimination claims, or indeed to discrimination claims generally.
102. Elias LJ, as I have said, conducted his analysis in terms of the particular disadvantage alleged to have been suffered by female discrimination claimants. But the label which he attached to this variant might suggest that one way of putting the indirect discrimination case would be that all persons bringing discrimination claims should be treated as a single composite class of people with – or claiming in respect of – protected characteristics. Although that is not the way that it was put, I do not believe that it would afford any escape from the problem identified by Elias LJ.
103. This variant depended on what Ms Monaghan alleged was a difference between the proportion of type B claimants who were women and the proportion of women in the workforce: the former was said to be 54% (cf. para. 85 above) and the latter 47%. Those figures, which were derived from an analysis of the 2008 Survey of Employment Tribunal Applications ("SETA"), were put before the Court in Unison 1 and accepted by it (see para. 78 of the judgment), subject to its ultimate refusal to reach any definitive conclusion. But in Unison 2 the Lord Chancellor put in evidence an analysis based on the 2013 SETA exercise, showing that the proportion of type B claimants who were women was 45%, i.e. broadly equivalent to the proportion of women in the workforce.
104. Elias LJ rejected criticisms advanced by Unison of the reliability of the newer figures adduced by the Lord Chancellor. He said, at para. 78:
"The claimant and the Commission take issue with the reliability of these new statistics. They say that they are based on a limited sample and the methodology is far from clear. I do not accept that. The Lord Chancellor has explained how the statistics were obtained. They were the result of a survey commissioned by the Department of Business, Innovation and Skill where almost two thousand claimants, selected randomly, were interviewed. The gender figures were based on the gender as reported by the claimants. The survey was completed before the introduction of the fees, but it was possible to identify what fee would now be payable from the information given. The results are subject to sampling error, but it seems to me that they are as reliable as any of the statistics available."
At para. 79 he recorded, though he did not explicitly accept, criticisms made by the Lord Chancellor of Unison's figures. He continued:
"80. It seems to me that all the figures which have been canvassed in these proceedings are to a greater or lesser extent unreliable. They are all derived from historical data, namely the periodic Surveys of Employment Tribunal Applications (SETA surveys).
81. Since the onus is on the claimant to show that there has been discrimination, I am not satisfied that the burden has been discharged here. Even if it has, the extent of any adverse impact is very small. That is relevant to the issue of justification."
105. Elias LJ's conclusion that Unison had not established any discrimination – that is, any disparate impact of the kind alleged – meant that he did not need to consider the issue of justification in this context. However, as I have already noted (see para. 88 above), he did proceed in the following paragraphs to consider the justifiability of any such disparate impact as Unison might arguably have shown.
106. Ms Monaghan contended that the Divisional Court in Unison 2 was wrong to find that Unison had failed to discharge the burden of proving disparate impact in this regard and that it should have accepted the contrary conclusion in Unison 1. We were shown the Lord Chancellor's evidence which Elias LJ accepted. This was in the form of a witness statement from Mr Jason Latham, the Deputy Director of HMCTS, served about ten days before the hearing, which set out the figures in question and explained that they were derived from the most recent SETA surveys. Unison asked for more details. These were given in a note (with attached spreadsheet) from Ms Chan, dated four days before the hearing, which also contained a minor correction. Unison's case is that it had had no real opportunity to test the accuracy of the statistics and that it was "wrong and/or perverse" of the Court to give them any weight.
107. I do not accept that argument. There could be no objection to the Lord Chancellor using the SETA figures as such, which had (in their 2008 version) been the basis of Unison's own case and with whose characteristics it and its lawyers would have been very familiar. Unison had been shown the detailed workings based on the 2013 figures that led to the 45% figure. It is no doubt unfortunate that those figures were produced so close to the start of the hearing (though we do not know how recently they had become available), but even if – which is not in fact asserted – Unison had made any formal objection to their admission it would have been well within the case management discretion of the Court to admit them while taking into account the limited opportunity that there had been to consider them. It is not suggested that any application was made to cross-examine Mr Latham or for an adjournment. It is also important to observe that although Elias LJ rejected Unison's challenge to the reliability of the newer figures his ultimate decision was based on a conclusion that the figures overall did not justify a finding of disparate impact. That was a finding that was plainly open to the Court.
108. That conclusion, taken with my conclusion as regards variant (2), means that I need not review the general consideration of the justification issue at paras. 82-91 of Elias LJ's judgment. I have in fact considered some of the points made by Ms Monaghan in connection with variant (1), but in so far as she made some additional points I need not further lengthen this judgment by addressing them here. I will only say that I was not persuaded that any flaw in his reasoning was shown.
CONCLUSION ON INDIRECT DISCRIMINATION
109. I would uphold the rejection by the Court in Unison 2 of the indirect discrimination challenge. That means that even if in Unison 1 the Court should have proceeded to a decision on this issue the appeal in that case must, as regards this issue, fail also.
(C) PUBLIC SECTOR EQUALITY DUTY
110. Section 149 of the 2010 Act provides, so far as material, as follows:
"(1) A public authority must, in the exercise of its functions, have due regard to the need to–
(a) eliminate discrimination, harassment, victimisation and any other conduct that is prohibited by or under this Act;
(b) advance equality of opportunity between persons who share a relevant protected characteristic and persons who do not share it;
(c) foster good relations between persons who share a relevant protected characteristic and persons who do not share it.
(2) A person who is not a public authority but who exercises public functions must, in the exercise of those functions, have due regard to the matters mentioned in subsection (1).
(3) Having due regard to the need to advance equality of opportunity between persons who share a relevant protected characteristic and persons who do not share it involves having due regard, in particular, to the need to -
(a) remove or minimise disadvantages suffered by persons who share a relevant protected characteristic that are connected to that characteristic;
(b) take steps to meet the needs of persons who share a relevant protected characteristic that are different from the needs of persons who do not share it;
(c) encourage persons who share a relevant protected characteristic to participate in public life or in any other activity in which participation by such persons is disproportionately low.
Sub-section (7) sets out the familiar protected characteristics: I need not list them.
111. There is a good deal of by now very familiar case-law about the application of section 149, and I will not recapitulate it here. Ms Monaghan referred us to the useful summary given by McCombe LJ at para. 26 of his judgment in Bracking v Secretary of State for Work and Pensions  EWCA Civ 1345,  EqLR 60.
112. At para. 60 of its judgment in Unison 1 the Divisional Court recorded the following:
"To comply with his duty under s.149, the Lord Chancellor has taken the following steps. First, there was a pre-consultation equality impact assessment dated 14 December 2011. Second, between 14 December 2011 and 6 March 2012 the Ministry of Justice undertook a public consultation on different fee-charging structures. The consultation paper was issued with 32 general questions and 12 equality impact questions. The consultation, at this stage, did not invite consideration of whether any fees should be introduced. Third, an impact assessment was issued dated 30 May 2012. Fourth, seven events were held at which 90 people from 60 organisations attended, and had the opportunity to put forward their views. Fifth, 140 written responses were received from different groups including employees, legal groups, business groups, advisory and equality groups, and from other interested parties. A detailed analysis and response to the consultation was published by the Ministry of Justice on 13 July 2012. This included consideration of potential discriminatory impacts and proposed adjustments to be made. Sixth, an equality impact assessment on charging fees in Employment and Employment Appeal Tribunals was prepared, dated 13 July 2012. Seventh, there was consultation on the new Civil Fee Remission Scheme undertaken between 18 April 2013 and 15 May 2013. The Ministry of Justice's consultation response was dated 9 September 2013, with an impact assessment. This included an equality statement containing an assessment of the new remission scheme upon different groups."
113. It is unnecessary to summarise the contents of each of the various documents there identified. But I should, I think, set out most of the summary of the conclusions of the Equality Impact Assessment ("the EIA") dated 13 July 2012 (i.e. the sixth of the steps listed by the Court). This reads12:
" This EIA considers the introduction of fees in employment tribunals (ET) and Employment Appeal Tribunal (EAT).
 In accordance with our duty under section 149 of the Equality Act 2010, we have considered the equality impacts of the proposed fees structure. The following is a summary of our overall assessment. Further detail of the impacts is given in the analysis sections below.
A. Prohibited Conduct
 The introduction of the proposed fee structure would not directly discriminate against people with a protected characteristic, because the fee changes would apply to all people irrespective of any protected characteristic; i.e. there is no less favourable treatment because of a protected characteristic.
 Many of those who responded were concerned that our proposals would put certain groups of people with a relevant protected characteristic at a particular disadvantage. We have carefully considered the evidence base for these assertions. All claimants, including those with protected characteristics, will have to pay a fee to use the ET and the EAT. We cannot rule out that this may have a greater impact on some people with particular protected characteristics than those who do not share that characteristic.
 Our analysis suggests that BME groups, women, younger people and disabled people are more likely to fall into the lower income brackets and therefore these groups would be more likely to qualify for partial or full fee remissions. For these groups it is unlikely that they will experience any particular disadvantage as the fee remission mitigation is considered likely to lessen the impact.
 Our analysis also suggests that mid to higher earners may experience the greatest negative impacts of the new fees and these people are more likely those aged 25 and over without children and people from a White ethnic group. These people are unlikely to qualify for full or partial remissions so the fee remission mitigation will not lessen the impacts. Other mitigations such as the free ACAS conciliation, not introducing a 3rd tier of fees and the ability to seek an order for the respondent to pay their fees should they win, are also considered to lessen the impacts. If people with these protected characteristics are unable to settle their ET issues via ACAS, they will have to pay fees up front and we are currently unclear whether or not this will deter claimants. Overall we do not consider that the introduction of the proposed fee structures is likely to amount to indirect discrimination under the Equality Act 2010.
Discrimination arising from disability and duty to make reasonable adjustments
Harassment and victimisation
B. Advancing equality of opportunity
 We consider that it is possible that the fee structure could impact on the duty to advance equality of opportunity if potential claimants with protected characteristics are put off from taking forward discrimination cases due to the introduction of fees. However, we think the mitigations we have proposed will protect access to justice for those with protected characteristics.
C. Fostering good relations
 We consider that to the extent that this obligation to the fee proposal is relevant it is reasonable to assume that the impact of fees on fostering good relations is neutral.
 In light of the responses, we have considered the impact of the fee proposals against the statutory obligations imposed by section 149 of the Equality Act 2010. These considerations have influenced our decision to recommend two fee levels instead of the original proposal to have three as mitigation against any potential indirect discrimination.
 Our assessment is that, based on the limited information available on the protected characteristics of individuals and their level of income, the introduction of fees will impact differently depending on the varying income profiles. We believe that the measures we have put in place would mitigate any equality impacts.
 For those who can afford to pay fees, the further mitigations are the power for the tribunal to order reimbursement if they are successful and the setting of fees below full cost recovery For those who cannot afford to pay fees, including those who can only make a contribution to the fee, the further mitigation is the availability of the remissions scheme to protect their access to the tribunal. The availability of Acas conciliation lessons the impacts for all irrespective of their financial position because if offers an alternative to making a claim.
 We therefore believe that with the existing mitigations and the additional measure of introducing two fee levels instead of three we can ensure that the proposals to introduce fees will mitigate the equality impacts.
 As part of the implementation process we consider how to ensure the fee payment system and forms are accessible by those who seek to use them and that information about the availability of remissions is widely accessible."
This is of course only the summary section: the full document runs to 32 pages (excluding annexes).
114. At paras. 63-68 of its judgment the Divisional Court considered various particular criticisms made by Unison of the Lord Chancellor's approach to his duties under section 149. It acknowledged force in some of them; but it concluded, at para. 69:
"The essential challenge now advanced relates to the Lord Chancellor's comment in the equality impact assessment that neither the Lord Chancellor, nor the Ministry of Justice, nor any of the respondents "can predict with any certainty what the impact of the introduction of fees will be". Of course, that does not justify a failure to consider the likely impact. But that impact was fully considered, even though its conclusion displeased the objectors. Again, as in relation to the first ground, we underline that the duty continues and the Lord Chancellor is under an obligation to assess the impact of the fee regime on the basis of evidence revealed in practice. If it turns out, as the objectors feared, that the introduction of fees has a damaging effect on the fundamental obligation of the Lord Chancellor and government to eliminate, so far as humanly possible, discrimination against those with relevant protected characteristics and advance equality of opportunity, then the Lord Chancellor will have to take such steps as are necessary by adjusting the regime. We acknowledge the genuine fear that the introduction of the fee regime will impede the vital goal of eliminating discrimination and advancing equality of opportunity. Whether that fear is well-founded may well depend on evidence yet to be obtained, as to how the regime has worked in practice. For the reasons we have given, we dismiss this ground of complaint."
115. Before us Unison advanced a large number of particular challenges to the adequacy of the Lord Chancellor's compliance with his duty under section 149 and/or to the reasoning of the Divisional Court in rejecting those challenges. I have sought without success to find some feature or features common to all or some of those criticisms; and I fear that I must simply consider them piecemeal. I will proceed by reference to a "speaking note" on the section 149 issue – largely, in fact, reproducing her skeleton argument – which Ms Monaghan helpfully put in during the hearing.
116. I should, however, say by way of preliminary that some of Ms Monaghan's criticisms seemed to me to fall into the error identified by Davis LJ in R (Bailey) v London Borough of Brent  EWCA Civ 1586 of approaching an EIA as if it were a forensic document (see at para. 102 of his judgment). An EIA is a working tool designed to ensure that decision-makers pay due regard to (as a shorthand) the equality impact of their decisions, and to act as a record that they have done so or at least that those impacts have been drawn to their attention. It will not typically be drafted by lawyers, nor typically should it be. To the extent that views are expressed on matters requiring assessment or evaluation the Court should go no further in its review than to identify whether the essential questions have been conscientiously considered and that any conclusions reached are not irrational. Inessential errors or misjudgments cannot constitute or evidence a breach of the duty.
117. Ms Monaghan's first point concerned the view expressed at para.  of the summary that the greatest negative impact would be on mid-to-high earners, who were more likely to be white and over-25, because they would not qualify for remission. We were not asked to examine the detailed basis for that conclusion, but on its face I can see nothing perverse about it. I think, however, that Ms Monaghan's real point was that the EIA fails to identify the likely discriminatory impact on women. She points out that the body of the report contains the (rather obvious) observation that the majority of sex discrimination and equal pay claims are brought by women and notes that "it could be argued that the introduction of fees will potentially have a differential impact on those women who claim on these grounds". But this is unfair to the reasoning in the summary. Para.  expressly acknowledges that the introduction of fees may have a greater impact on people with particular protected characteristics, but para.  concludes that this impact is likely, in effect, to be cancelled out for (among others) women by the availability of remission. The argument may be right or wrong, but the EIA cannot be said to ignore the possibility of a differential impact on women.
118. Secondly, Ms Monaghan said that in so far as the EIA relies on the availability of fee remission "they get this wholly wrong", because the Lord Chancellor's prediction was that 8.6% of claimants would benefit from remission whereas it has since turned out that only 3.9% do so. But the fact that a particular prediction turns out wrong is not by itself a sufficient basis for finding a breach of the duty under section 149. (Nor in any event would the fact that take-up of remission was less than expected mean that there was indirect discrimination.) Ms Monaghan commented that the error might reflect the fact that there was no separate EIA of the new remissions scheme as it applied to ETs; but she made it clear in her oral submissions that that was not in itself relied on as a separate breach13.
119. Thirdly, Ms Monaghan pointed out that the Lord Chancellor did not enter into any consultation about the principle of charging fees in the Tribunals but only about how such a scheme should work. She said that the duty under section 149 applied to both. That is no doubt true in theory but I cannot see its relevance in practice. The Lord Chancellor had to consider the equality impacts of the particular scheme which he was considering introducing. That is what, on the face of it, he did.
120. Fourthly, Ms Monaghan notes and adopts a finding by the Divisional Court that there is no mention in the EIA – or, so far as it could see, in any of the associated materials – of the level of awards relative to fees (cf. para. 45 above) or of the difficulties of recovery. But I do not accept that that omission is of such a fundamental character as to vitiate the assessment.
121. Fifthly, it is said that "no evidence was gathered as to the likely deterrent effect of the introduction of fees … on different protected groups". It is not clear to me, and Ms Monaghan did not explain in her submissions, what kind of evidence could or should reasonably have been sought. It is a matter of common sense that the introduction of a fee for a service14 which has previously been available for free is likely to produce some reduction in take-up, even from those who can realistically afford to pay; but the extent of that reduction must be very difficult to predict, and asking people in advance on a hypothetical basis would seem unlikely to produce reliable data. I can see nothing wrong in making a reasonable judgement and then monitoring the outcome with a view to making any adjustments that may seem necessary: the section 149 duty is ongoing. Elias LJ made a similar point, albeit that the context was different (he was not of course concerned with the section 149 claim – see para. 29 (4) above), in the course of his discussion of justification in his judgment in Unison 2. In relation to the Lord Chancellor's aim that the introduction of fees should encourage early settlement of claims, he said, at para. 86:
"The Commission advanced … [an] argument to the effect that there is no hard evidence to support the Lord Chancellor's assessment but as the EAT pointed out in Constable of West Yorkshire Police and another v Homer  ICR 223 para. 48, concrete evidence is not always required. A reasonable and rational view about what effects a particular policy is likely to have will in principle suffice to justify its adoption, although the impact of the policy will have to be kept under consideration to ensure that the ends justify the means. Were it otherwise, government would be stifled in its ability to introduce new and untried measures because of the uncertainty of their impact."
I respectfully agree with that.
122. Sixthly, Ms Monaghan contended that (to quote from her skeleton argument)
"the Households Below Average Income survey used by the EIA to address household income and living standards is predicated on the generalised assumption that both partners in a couple benefit equally from household income; but women in fact are often not the beneficiaries of an equal share in household income"
and she referred in her speaking note to an acknowledgment in the Lord Chancellor's evidence that household income is not always equally shared. I cannot accept that this, as she describes it, "unfounded assumption" would be capable of vitiating the EIA, even in combination with the other errors alleged. I refer to what I say at para. 116 above.
123. Seventhly, she said that the Lord Chancellor failed to ask whether the fees regime would adversely impact on the need to eliminate discrimination or on the need to advance equality, as identified in section 149 (1) (a) and (b), by making it harder to bring discrimination claims in the ET. But that question is specifically addressed at para.  of the EIA. I think that her real criticism, which she makes explicitly elsewhere, is that the point apparently being made in the EIA is that the availability of remission meant that there would be no inhibition on the willingness of potential claimants to bring claims, and that that was obviously wrong: even if the level of fees were otherwise unobjectionable, the mere fact of having to pay a fee would be bound to have some deterrent effect. But I do not read para.  as saying that no reduction in the number of claims was expected; and if it did it would be inconsistent with the aim, expressed elsewhere in the document, of encouraging resolution of potential claims by settlement. In my view it is reasonably clear that the assessment in the EIA, rightly or wrongly, was that the new regime would not deter claimants with good, or in any event arguable, claims from bringing them to the Tribunal if they could not be resolved otherwise. The figures discussed earlier in this judgment do indeed suggest that that was wrong and that the introduction of fees has deterred, even if it has not positively prevented, many claimants from bringing discrimination (and other) claims; and it is reasonable to assume that some at least of those claims will have been good15 and have not been resolved by settlement. It may be that the belief that it would ever be otherwise was over-optimistic. But even if that were so, it would not be a reason for finding a breach of the section 149 duty: the Lord Chancellor had due regard to the issue, even if events have proved him wrong. No doubt the Lord Chancellor will be considering in the current review the extent to which the deterrent effect of the fees regime adversely affects the needs identified at heads (a) and (b) under section 149 (1).
124. Finally, Ms Monaghan referred to the phrase in para.  of the EIA "to the extent that this obligation [sic] to the fee proposal is relevant". She said that that suggests that the Lord Chancellor questions whether a deterrent effect on the bringing of discrimination claims is relevant to the need to foster good relations identified in section 149 (1) (c). The meaning of para.  is frankly rather opaque, but I will assume that Ms Monaghan's reading of it is correct. She submitted that that doubt is unjustifiable: she did not quite spell out why, but I take the thinking to be that the bringing of successful (or perhaps even unsuccessful) discrimination claims raises awareness of equality issues and so promotes behaviour which respects diversity and so fosters good relations between the groups in question. I am not sure that I agree that the draftsman had that kind of indirect connection in mind. The various sub-heads of section 149 (1) ought, so far as possible, be given separate work to do, and I think that inhibitions on the ability to bring discrimination claims are amply covered by (a) and/or (b), leaving (c) to address other matters. But we do not need to decide this point. The criticism of substance is that the EIA did not anticipate the deterrent effect of fees on the bringing of discrimination claims, and I have already addressed that.
125. I would accordingly dismiss the appeal against the rejection by the Divisional Court in Unison 1 of the public sector equality duty challenge.
WAS THE CHALLENGE IN UNISON 1 PREMATURE ?
126. I should start by setting out what the Divisional Court said about prematurity. Paras. 89-90 of the judgment read (so far as relevant):
"89. This brings us to a fundamental difficulty with the whole of this case. Brought as it was in the belief that the lawfulness of the regime had to be challenged as a matter of urgency, and in any event within three months, the court has been faced with judging the regime without sufficient evidence, and based only on the predictions of the rival parties throughout and after the hearing. Parliament decided, by affirmative resolution, to introduce the regime, authorised by statute, and debated and positively affirmed by both Houses of Parliament. Quite apart from the continuing obligation to fulfil the duties identified in the 2010 Act, the defendant has himself undertaken to keep the issue of the impact of this regime under review. If it turns out that over the ensuing months the fees regime as introduced is having a disparate effect on those falling within a protected class, the defendant would be under a duty to take remedial measures to remove that disparate effect and cannot deny that obligation on the basis that challenges come too late. It seems to us more satisfactory to wait and see and hold the defendant to account should his optimism as to the fairness of this regime prove unfounded. We believe both the claimant and the intervener will be, and certainly should be, astute to ensure that accurate figures and evidence are obtained as to the effect of this regime.
90. No doubt the defendant will also be doing the same, if he is successfully to resist a future challenge. In the meantime, we think that the fundamental flaw in these proceedings is that they are premature and that the evidence at this stage lacks that robustness necessary to overturn the regime. …"
127. My conclusions on the substantive issues mean that it is not necessary to decide whether that approach was right, but I will state my views briefly in case a similar situation arises in the future. There was of course nothing wrong in Unison seeking to strike down the Order prior to its implementation on the basis of procedural failures or of unlawful effects which it would necessarily have if implemented: on the contrary, the sooner any such challenge was raised the better. But the position is not so straightforward in so far as the challenge was based on predictions as to the effect of the Order which could not definitively be established from the evidence available pre-implementation. Even in such a case Unison could in principle have succeeded by showing that the risk of an unlawful impact was so great that the Lord Chancellor could not rationally have discounted it, and I do not therefore believe that the proceedings in Unison 1 were formally premature. But I think the Court was right to be reluctant to reach a decision on the effectiveness and discrimination challenges on the material before it. It is a strong thing to strike down legislation on the basis of disputed predictions as to its effect when the passage of a comparatively short period of time will prove their correctness or otherwise. In my view it was a proper exercise of the Court's discretion in the present case, given the real difficulties with the quality of the evidence available pre-implementation, to decline to grant any relief.
128. I would dismiss both appeals.
Lord Justice Davis:
129. I agree.
Lord Justice Moore-Bick:
130. I also agree.
ANNEX 1: "TABLE 2" (omitting the column identifying the provisions conferring jurisdiction)
1. Application by the Secretary of State to prohibit a person (Section 3A EAA)
2. Application by a person subject to a prohibition order to vary or set it aside (Section 3C EAA)
3. Appeal against improvement or prohibition notice (Section 24 HSWA)
4. Appeal against assessment of training levy (Section 12 ITA)
5. Complaint of deduction of unauthorised subscriptions (Section 68 TULR(C)A)
6. Complaint relating to failure to deduct or refuse to deduct amount to a political fund (Section 86 TULR(C)A)
7. Complaint that an employer has failed to permit time off for carrying out trade union duties (Section 168)
8. Complaint that an employer has failed to permit time off for union learning representatives (Section 168A)
9. Complaint that an employer has failed to pay for the time off for union learning representatives (Section 169 TULR(C)A)
10. Complaint that an employer has failed to permit time off for trade union activities (Section 170 TULR(C)A)
11. Complaint that employer has failed, wholly or in part, to pay remuneration under a protective award (Section 190 TULR(C)A)
12. Complaint that the Secretary of State has not paid, or has paid less than, the amount of relevant contributions which should have been paid into a pension scheme (Section 124 PSA)
13. Breach of contract, except where the employer's contract claim is made by way of application as part of the employer's response to the employee's contract claim (as to which, see instead article 4 and Schedule 1 to this Order)
14. Reference to determine what particulars ought to be included in a statement of employment particulars or changes to particulars (Sections 1 and 4 ERA)
15. Reference to determine what particulars ought to be included in an itemised pay statement (Section 8 ERA)
16. Complaint of unauthorised deductions from wages (Section 13 ERA)
17. Complaint that employer has received unauthorised payments (Section 15 ERA)
18. Complaint that employer has failed to pay guaranteed payment (Section 28 ERA)
19. Complaint that employer has failed to permit time off for public duties (Section 50 ERA)
20. Complaint that employer has refused to permit, or has failed to pay for, time off to look for work or arrange training (Sections 52 and 53 ERA)
21. Complaint that employer has refused to allow, or has failed to pay for, time off for ante-natal care (Sections 55, 56, 57ZA and 57ZB ERA)
22. Complaint that employer has refused to allow time off for dependants (Section 57A ERA)
23. Complaint that employer has failed to allow, or to pay for, time off for trustee of pension scheme (Sections 58 and 59 ERA)
24. Complaint that employer has failed to allow, or to pay For, time off for employee representative (Sections 61 and 62 ERA)
25. Complaint that employer has failed to allow, or to pay for, time off for young people in Wales and Scotland (Sections 63A and 63B ERA)
26. Complaint that employer has failed to pay for time off on medical or maternity grounds (Sections 64, 68 and 68C)
27. Complaint that employer has failed to deal with an application in relation to study or training in accordance with regulations or refused the application on the basis of incorrect facts (Sections 63D to 63H ERA)
28. Complaint that employer has unreasonably failed to provide a written statement of reasons for dismissal or the particulars are inadequate or untrue (Section 92 ERA)
29. Reference in respect of a right to redundancy payment (Section 135 ERA)
30. Reference related to payment out of National Insurance Fund (Section 166 ERA)
31. References related to payments equivalent to redundancy payments (Sections 167, 168 and 177 ERA)
32. Complaint that the Secretary of State has failed to make any, or insufficient, payment out of the National Insurance Fund (Section 182 ERA)
33. Appeal against a notice of underpayment (Section 19C NMWA)
34. Appeal against a notice issued by the Commission for equality and Human Rights where the notice relates to an unlawful act (Section 21EA 2006)
35. Complaint that prospective employer made enquiries about disability or health (Section 60EA 2010)
36. Application in relation to the effect of a non-discrimination rule in an occupational pension scheme (Section 61EA 2010)
37. Complaint in relation to a breach of a maternity equality clause (Section 73EA 2010)
38. Complaint in relation to a breach of, or application in relation to the effect of, a maternity equality rule in an occupational pension scheme (Section 75EA 2010)
39. Complaint in relation to terms prohibiting discussions about pay (Section 77EA 2010)
40. Complaint that a term in a collective agreement is void or unenforceable (Section 145EA 2010)
41. Appeal of decisions of compensating authority (Regulation 42 CEC)
42. Complaint that employer has failed to pay for remuneration time off for safety representative (Regulation 4(2) of, and Schedule 2 to, the SRSC)
43. Reference that there has been a failure to consult with employee representatives about contracting out of pension scheme (Regulation 4 OPS(CO)R and regulation 9 of OPS (DI)R)
44. Complaint that employer has failed to pay for time off to carry out Safety Representative duties or undertake training (Regulation 7 of, and Schedule 1 to, the HSCE)
45. Complaint that employer has refused to allow annual leave or make payment in respect of annual leave (Regulations 13, 13A,14 and 16 WTR)
46. Appeal against improvement or prohibition notice (Paragraph 6 of Schedule 3 to WTR)
47. Appeal against improvement or prohibition notice (Regulation 23 COMAH)
48. Complaint in relation to refusal of annual leave or to make payment (Regulation 11 MSR)
49. Complaint in relation to refusal to provide paid annual leave (Regulation 4 CAR)
50. Complaint in relation to failure to provide free health assessments (Regulation 5 CAR)
51. Complaint in relation to refusal of annual leave or to make payment (Regulation 11 FVR)
52. Complaint that employer has refused to allow or failed to pay for time off for information and consultation or negotiating representatives (Regulations 27 and 28 ICR)
53. Appeal against improvement notice (Paragraph 6(2) of Schedule 2 to the RTR)
54. Complaint in relation to failure to pay compensation in pursuance of tribunal order (Regulation 15(7) and 15(8) TUPE)
55. Complaint that employer has failed to allow, or pay for, time off for functions as employee representative (Paragraphs 2 and 3 of the Schedule to OPR)
56. Complaint that employer has failed to allow, or pay for, time off for members of special negotiating body (Regulations 28 and 29 EC SR)
57. Complaint that employer has failed to allow, or pay for, time off for members of special negotiating body (Regulations 43 and )
58. Appeal against notice from Health and Safety Executive or a local authority Part 2 of (Regulation 21 and Schedule 8 to REACHER)
59. Reference to determine what particulars ought to be Included in an itemised statement of stipend (Regulation 6 EOR)
60. Reference to determine what particulars ought to be included in a statement of particulars or changes to particulars (Regulations 3 and 6 to EOR)
61. Complaint that employer has failed to allow, or pay for, time off for members of special negotiating body (Regulations 26 and 27 ELLR)
The abbreviations are as follows:
CAR: Civil Aviation (Working Time) Regulations 2004
CCBR: Companies (Cross-Border Mergers) Regulations 2007
CEC: Colleges of Education (Compensation) Regulations 1975
COMAH: Control of Major Accident Hazards Regulations 2015
EA 2006: Equality Act 2006
EA 2010: Equality Act 2010
EAA: Employment Agencies Act 1973
ECSR: European Cooperative Society (Involvement of Employees) Regulations 2006
EJOs: Employment Tribunals Extension of Jurisdiction (England and Wales) Order 1994 and Employment Tribunals Extension of Jurisdiction (Scotland) Order 1994
EOR: Ecclesiastical Offices (Terms of Service) Regulations 2009
ELLR: European Public Limited-Liability Company (Employee Involvement) (Great Britain) Regulations 2009
ERA: Employment Rights Act 1996
ETA: Employment Tribunals Act 1996
FVR: Fishing Vessels (Working Time: Sea-fishermen) Regulations 2004
HSCE: Health and Safety (Consultation with Employees) Regulations 1996
HSWA: Health and Safety at Work etc Act 1974
ICR: Information and Consultation of Employees Regulations 2004
ITA: Industrial Training Act 1982
MSR: Merchant Shipping (Working Time: Inland Waterways) Regulations 2003
NMWA: National Minimum Wage Act 1998
OPR: Occupational and Personal Pension Schemes (Consultation by Employers and Miscellaneous Amendment) Regulations 2006
OPS(CO)R: Occupational Pension Schemes (Contracting-Out) Regulations 1996
OPS(DI)R: Occupational Pensions Schemes (Disclosure of Information) Regulations 1996
PSA: Pension Schemes Act 1993
REACHER: REACH Enforcement Regulations 2008
RTR: Road Transport (Working Time) Regulations 2005
SRSC: Safety Representatives and Safety Committees Regulations 1977
TULR(C)A: Trade Union and Labour Relations (Consolidation) Act 1992
TUPE: Transfer of Undertakings (Protection of Employment) Regulations 2006
WTR: Working Time Regulations 1998
ANNEX 2: EU-DERIVED RIGHTS IN THE ET
Paid time off for health and safety representatives
Health and Safety (Consultation with Employees) Regulations 1996
Written statement of terms and conditions
ERA 1996, Part I
Maternity rights – right to suitable alternative work or pay if suspended
ERA 1996, Part VII
Maternity/parental rights – maternity/parental leave
Directives 92/85/EEC and 2010/18/EC
ERA 1996, Part VI, VIII; Maternity and Parental Leave Regulations 1999
Maternity/parental rights – prohibition of detriment
Directives 92/85/EEC and 2010/18/EC
Maternity and Parental Leave Regulations 1999, reg 19; ERA 1996, Part V
Maternity/parental rights – unfair dismissal
Directive 92/85/EEC and 2010/18/EC
Maternity and Parental Leave Regulations 1999, reg 20; ERA 1996, Part X
Protection of young people at work
Part-time workers discrimination
Part-Time Workers (Prevention of Less Favourable Treatment) Regulations 2000
Part-time workers – unfair dismissal
Part-Time Workers (Prevention of Less Favourable Treatment) Regulations 2000, reg 7; ERA 1996, Part X
Information and consultation on collective redundancies; enforcement of protective award
TULR(C)A 1992, Part IV, Chapter II
Fixed-term workers discrimination
Fixed-Term Employees (Prevention of Less Favourable Treatment) Regulations 2002
Fixed-term workers – prevention of abuse (employee treated as permanent employee)
Fixed-Term Employees (Prevention of Less Favourable Treatment) Regulations 2002, regs 8-9
Discrimination (etc) related to sex, race, disability, sexual orientation, religion or belief, and age
Articles 8, 10, and 157; Directives 2000/43/EC, 2000/78/EC, 2006/54/EC
EA 2010, Part 5
Articles 8, 10, and 157; Directive 2004/56/EC
EA 2010, Parts Part 5, Chapter 3
Transfer of undertaking –information and consultation
Transfer of undertaking – unfair dismissal
TUPE 2006, reg 7; ERA 1996, Part X
Information and consultation of employees in general – paid time off
Information and Consultation of Employees Regulations 2004
Information and consultation of employees – prohibition of detriment
Information and Consultation of Employees Regulations 2004; ERA, sections 48-49
Information and consultation of employees – unfair dismissal
Information and Consultation of Employees Regulations 2004; ERA, Part X
Working time – daily rest, weekly rest, rest breaks, paid annual leave
Working time – prohibition of detriment
WTR 1998; ERA 1996, Part V
Protection in case of employer's insolvency (payments from Secretary of State)
ERA 1996, sections 166-170, Part XII
Agency workers discrimination
Agency Workers Regulations 2010
Agency workers – unfair dismissal
Agency Workers Regulations 2010, reg 17; ERA 1996, Part X
1 In the Order as originally made equal pay claims were listed as type A; but that was an error and they were subsequently removed and so became type B – see the Tribunals Fees (Miscellaneous Amendments) Order 2014.
2 From now on I will use the shorthand "claimant" to refer also to appellants in the EAT. This is not strictly accurate, since appellants may well have been respondents in the ET. But it will be rare for such respondents to be individuals.
3 Ms Monaghan in her skeleton argument provided a helpful list of EU-derived claims, which I annex. (Here and elsewhere I refer to the argument as Ms Monaghan's alone for convenience and not with any intention to devalue Mr Purchase's contribution.)
4 As regards the UK, Elias LJ referred at para. 24 of his judgment in Unison 2, to Ahmed v HM Treasury  UKSC2,  2 AC 534, per Lord Phillips at para. 146 (p. 654 B-C).
5 Similar arguments would, it seems, apply also to County Court claims, primarily by reference to article 6 of the Convention (although the County Court does have jurisdiction over some EU-derived claims). The fees system is different, being related to the value of the claim, but it can produce broadly similar figures (see Unison 1, at para. 52); and, as I have said, the remission regime is the same.
6 See, e.g., Magorrian v Eastern Health and Social Services Board (C-246/96),  ICR 979, at para. 37 (p. 1003).
7 The skeleton arguments refer to two other Strasbourg cases of the same character – Weissman v Romania (63945/00) and Kijewska v Poland (73002/01) – but we were not taken to them and they contain no statements of principle not apparent from the three cases to which I refer.
8 Mr Barr pointed out that average awards are much higher – very approximately double; but that reflects the impact of comparatively few very large awards and is less significant than the median in this context.
9 There may be an analogy – though I do not suggest that it is exact – with the approach taken to the legislation relating to the automatic deportation of foreign criminals. In MF (Nigeria) v Secretary of State for the Home Department  EWCA Civ 1192,  1 WLR 544, this Court held that the provision for deportation not to proceed "in exceptional circumstances" could be interpreted broadly enough to cover all those cases in which deportation would involve a breach of article 8 of the Convention.
10 As, for example, Lord Walker noted in Secretary of State for Trade and Industry v Rutherford (no. 2)  UKHL 19,  ICR 785, at para. 49 (p. 799A).
11 See the Woodcock case referred to by the Supreme Court when it was before the EAT ( UKEAT 0489,  ICR 143), at para. 32 (pp. 155-6).
12 Unhelpfully, the original has no paragraph numbers, but I have inserted them for ease of cross-reference.
13 This short-circuits a dispute about an observation made by the Divisional Court at para. 68 of its judgment to the effect that "it can be said that there was no assessment at all in relation to the changes to the remission scheme". Mr Barr said that a "not" had evidently been omitted before "said". Ms Monaghan responded that Ms Chan had suggested just such a correction when the Court's judgment had been circulated in draft but that it had not been adopted. I am bound to say that I think that Mr Barr is probably right despite the Court's failure to make the correction, but the point can be left to die a quiet death.
14 I use the word "service" for convenience, but there are dangers in the use of this terminology in the context of the provision by the state of access to justice.
15 Ms Monaghan told us that the available statistics did not suggest that the success rate of claims brought since the introduction of the fees regime was significantly higher than before, so that there is no basis for an inference that it is the weaker claims that have been deterred. But we did not go into the point in any detail.