Tyne and Wear Passenger Transport Executive (trading as Nexus) v National Union of Rail, Maritime and Transport Workers & Anor.

The most recent authoritative consideration of the law of res judicata is in the decision of the Supreme Court in Virgin Atlantic Airways Ltd v Zodiac Seats UK Ltd [2013] UKSC 46, [2014] 1 AC 160. Lord Sumption delivered the main judgment. The principal issue was whether cause of action estoppel was absolute not only as regards points decided in the earlier proceedings but also as regards points potentially affecting the existence of the cause of action but which had not been raised in those proceedings and had accordingly not been the subject of any decision. The resolution of that issue required a thorough consideration of the relationship of not only cause of action estoppel but also issue estoppel with the principle in Henderson v Henderson (1843) 3 Hare 100.

At para. 17 of his judgment Lord Sumption gave a helpful overview of various principles applied by the courts which could be grouped together under the heading “res judicata”. He said (so far as relevant for our purposes):

“Res judicata is a portmanteau term which is used to describe a number of different legal principles with different juridical origins. As with other such expressions, the label tends to distract attention from the contents of the bottle. The first principle is that once a cause of action has been held to exist or not to exist, that outcome may not be challenged by either party in subsequent proceedings. This is ‘cause of action estoppel’. It is properly described as a form of estoppel precluding a party from challenging the same cause of action in subsequent proceedings. Secondly, there is the principle, which is not easily described as a species of estoppel, that where the claimant succeeded in the first action and does not challenge the outcome, he may not bring a second action on the same cause of action, for example to recover further damages … . Third, there is the doctrine of merger, which treats a cause of action as extinguished once judgment has been given upon it, and the claimant’s sole right as being a right upon the judgment. … Fourth, there is the principle that even where the cause of action is not the same in the later action as it was in the earlier one, some issue which is necessarily common to both was decided on the earlier occasion and is binding on the parties: Duchess of Kingston’s Case (1776) 20 St Tr 355. ‘Issue estoppel’ was the expression devised to describe this principle by Higgins J in Hoysted v Federal Commissioner of Taxation (1921) 29 CLR 537, 561 and adopted by Diplock LJ in Thoday v Thoday [1964] P 181, 197-198. Fifth, there is the principle first formulated by Wigram V-C in Henderson v Henderson (1843) 3 Hare 100, 115, which precludes a party from raising in subsequent proceedings matters which were not, but could and should have been raised in the earlier ones. Finally, there is the more general procedural rule against abusive proceedings, which may be regarded as the policy underlying all of the above principles with the possible exception of the doctrine of merger.”

(I have added the underlinings for ease of reference.) For present purposes we are concerned mainly with the first, fourth and fifth of those principles, and consistently with Lord Sumption’s approach I use the term res judicata to refer to them collectively.

In the following paragraphs of his judgment Lord Sumption reviews the inter-relation of those principles as they apply in cases where a party in later proceedings seeks to raise a point which could have been, but was not, raised in earlier proceedings between the same parties. He starts, at paras. 20-21, by examining the decision of the House of Lords in Arnold v National Westminster Bank plc [1991] 2 AC 93. Having analysed the reasoning in the speech of Lord Keith, with whom the rest of the committee agreed, at para. 22 of his judgment he summarises the position as follows:

“Arnold is … authority for the following propositions:

(1)

Cause of action estoppel is absolute in relation to all points which had to be and were decided in order to establish the existence or non-existence of a cause of action.

(2)

Cause of action estoppel also bars the raising in subsequent proceedings of points essential to the existence or non-existence of a cause of action which were not decided because they were not raised in the earlier proceedings, if they could with reasonable diligence and should in all the circumstances have been raised.

(3)

Except in special circumstances where this would cause injustice, issue estoppel bars the raising in subsequent proceedings of points which (i) were not raised in the earlier proceedings or (ii) were raised but unsuccessfully. If the relevant point was not raised, the bar will usually be absolute if it could with reasonable diligence and should in all the circumstances have been raised.”

(The italicisation in propositions (2) and (3) is mine.)

In the present case we are concerned with the application of res judicata to points not raised in the first proceedings. As to that, the essential points from Lord Sumption’s summary are:

At paras. 23-26 Lord Sumption considers and rejects an argument that the degree of flexibility recognised in his formulation was not available in the context of res judicata and was relevant only in the context of abuse of process. He reviews the case-law about the basis of the rule in Henderson v Henderson, in particular the decision of the House of Lords in Johnson v Gore-Wood & Co [2000] UKHL 65, [2002] 2 AC 1, and concludes, at para. 25, that res judicata and abuse of process “are distinct although overlapping legal principles with the common underlying purpose of limiting abusive and duplicative litigation”. That being so, in my analysis below I have in the interests of clarity dealt with the two principles separately, as the Judge also did; but, as will appear, the same substantive considerations are likely to be relevant in both contexts.

Res judicata only applies where the parties to the second proceedings are the same as the parties to the first proceedings or are their “privies”. Privies are traditionally treated as falling into three classes – “privies in blood”, “privies in law” and “privies in estate or interest”. In this case we need only consider privity of interest.

The first full examination in the authorities of the concept of privity of interest is in the judgment of Sir Robert Megarry V-C in Gleeson v J Wippell & Co [1977] 1 WLR 510, at pp. 514-517. I need not set the passage out in full. He begins by describing the concept as “protean, … and at times … almost capable of meaning all things to all men” (p. 514D). At p. 517 C-D he says:

“This is difficult territory: but I have to do the best I can in the absence of any clear statement of principle. First, I do not think that in the phrase ‘privity of interest’ the word ‘interest’ can be used in the sense of mere curiosity or concern. Many matters that are litigated are of concern to many other persons than the parties to the litigation, in that the result of a case will at least suggest that the position of others in like case is as good or as bad as, or better or worse than, they believed it to be. Furthermore, it is a commonplace for litigation to require decisions to be made about the propriety or otherwise of acts done by those who are not litigants. Many a witness feels aggrieved by a decision in a case to which he is no party without it being suggested that the decision is binding upon him.

Second, it seems to me that the substratum of the doctrine is that a man ought not to be allowed to litigate a second time what has already been decided between himself and the other party to the litigation. This is in the interest both of the successful party and of the public. But I cannot see that this provides any basis for a successful defendant to say that the successful defence is a bar to the plaintiff suing some third party, or for that third party to say that the successful defence prevents the plaintiff from suing him, unless there is a sufficient degree of identity between the successful defendant and the third party. I do not say that one must be the alter ego of the other: but it does seem to me that, having due regard to the subject matter of the dispute, there must be a sufficient degree of identification between the two to make it just to hold that the decision to which one was party should be binding in proceedings to which the other is party. It is in that sense that I would regard the phrase ‘privity of interest’.”

(He goes on – “thirdly” – to explore the effect in the case before him of the rule that “for privity with a party to the proceedings to take effect, it must take effect whether that party wins or loses”; but I need not summarise that part.)

In Johnson v Gore-Wood & Co [2000] UKHL 65, [2002] 2 AC 1, Lord Bingham approved a finding that the owner of a company which was his “corporate embodiment” could be prevented by issue estoppel from revisiting a decision made in proceedings brought by the company. He deprecated a “formulaic approach” and expressly approved the second paragraph in the passage which I have quoted from Gleeson: see p. 32 E-G.

The most recent review of the law in this area appears in the judgment of Floyd LJ in Resolution Chemicals Ltd v H Lundbeck A/S [2013] EWCA Civ 924, [2014] RPC 5: see paras. 22-35. After considering various authorities, including Gleeson, at para. 32 he summarises the correct approach as follows:

“… [I]n my judgment a court which has the task of assessing whether there is privity of interest between a new party and a party to previous proceedings needs to examine (a) the extent to which the new party had an interest in the subject matter of the previous action; (b) the extent to which the new party can be said to be, in reality, the party to the original proceedings by reason of his relationship with that party, and (c) against this background to ask whether it is just that the new party should be bound by the outcome of the previous litigation.”

The formulation of element (c) reflects the particular facts of Resolution, where the estoppel was prayed in aid against the new party. In the present case it is the new parties, the Unions, who seek to invoke the estoppel against the party common to both actions, Nexus; but what matters is that the ultimate question is whether it is just to allow the issue in question to be relitigated.

The process envisaged by that summary is one of evaluation rather than the application of hard-edged criteria: that is apparent not only from the broad terms of the formulation of the question at (c) but from the formulation of elements (a) and (b) in terms of degree – “the extent to which”. That is consistent with Sir Robert Megarry’s reference in Gleeson to “a sufficient degree” of identification between the two parties to “make it just” that a decision to which one was party should be binding in proceedings involving the other. I should also note that at para. 30 Floyd LJ observes that an estoppel would arise against a party to a second action raising an issue decided in an earlier action where “in effect, he represents the party in the first action” and refers to the example given by Lord Reid in Carl Zeiss Stiftung v Rayner & Keeler Ltd (no. 2) [1967] 1 AC 853 (see pp. 911-912) of a case where the first action is brought by a servant or agent of the claimant in the second.

As appears from what Lord Sumption says as quoted at para. 41 above, there is considerable overlap between the operation of res judicata and the power to strike out an action as an abuse of process because it raises issues which could have been raised in earlier litigation. As regards the latter, the correct approach is authoritatively established in pp. 30-31 of the speech of Lord Bingham in Johnson v Gore-Wood. The only point that I need note for present purposes is his statement that what is required is:

“… a broad, merits-based judgment which takes account of the public and private interests involved and also takes account of all the facts of the case, focusing attention on the crucial question whether, in all the circumstances, a party is misusing or abusing the process of the court by seeking to raise before it the issue which could have been raised before. As one cannot comprehensively list all possible forms of abuse, so one cannot formulate any hard and fast rule to determine whether, on given facts, abuse is to be found or not.”

A party can be precluded on this basis from raising an issue even where it was neither a party to the previous proceedings nor their privy. That was decided by this Court in Aldi Stores Group Ltd v WSP Group plc [2007] EWCA Civ 1260, [2008] 1 WLR 748. In that case the claimant sought to raise a claim against two defendants based on allegations which it had already raised against other defendants in earlier proceedings. The defendants contended that that was as an abuse of process because the claimant could and should have joined them in the original proceedings. The claimant argued that there could be no abuse of process unless they and the defendants in the original proceedings were privies. At para. 10 of his judgment (with which Longmore and Wall LJJ agreed) Thomas LJ rejected that argument. He said:

“The fact that the defendants to the original action and to this action are different is a powerful factor in the application of the broad-merits based judgment; it does not operate as a bar to the application of the principle.”

(The abuse argument was in fact dismissed on other grounds, but those are not material for our purposes.) To this extent abuse of process is available as an argument where res judicata would not be.

THE ANDERSON CLAIMANTS

As explained at para. 18 above, the Anderson proceedings could only formally be concerned with deductions prior to the presentation of the claim, i.e. those alleged to have taken place between the date that the Letter Agreement came into force (Footnote: ⁵), and 19 June 2015 (“pre-complaint deductions”). It follows that any decision in those proceedings could not give rise to a cause of action estoppel as regards subsequent deductions: if it were to support a defence of res judicata that could only be on the basis of issue estoppel. That means that I need to consider the issue of res judicata, as it affects the Anderson claimants, separately as regards (a) the pre-complaint deductions and (b) any post-complaint deductions. However, it will be convenient to deal first with an issue which is common to both groups, namely whether the mistake case “could with reasonable diligence and should in all the circumstances have been raised” – see Lord Sumption’s propositions (2) and (3).

Lord Sumption’s formulation refers both to whether the point in question “could” have been raised and to whether it “should” have been. There is a risk of being over-nice in an analysis which is ultimately directed to a broad question of justice, and the distinction between “could” and “should” may not always be clearly marked. However, the Judge and the parties addressed the two questions separately, and I think I should do the same.

I will start by stating my own conclusion, without reference to the parties’ submissions or the Judge’s reasoning, though I will return to them in due course.

In my view Nexus could have raised the mistake case in the Anderson proceedings. It is of course true that an ET exercising its jurisdiction under Part II of the 1996 Act has no power to order rectification as such; the only relief that it can grant is as specified in section 24 of the Act (see para. 15 above). However, the mistake case could have been deployed by way of defence without the need for an order for rectification. As explained in the following paragraphs, that is the case in the ordinary courts, and I see no reason why it should not equally be the case in an unlawful deductions claim in the ET.

The position in the courts is established by two decisions of the Common Pleas Division of the High Court following the enactment of the Supreme Court of Judicature Act 1873. Section 34 of the 1873 Act assigned claims for “the rectification, or setting aside, or cancellation of deeds or other written instruments” to the Chancery Division alone, but the question arose whether other divisions of the High Court could determine issues of that character where they arose by way of defence. In short, it was held that they could: as succinctly summarised at para. 16-003 of Snell, “any Division may give effect to a defence of rectification as regards past transactions without actually rectifying the instrument”. It is, however, necessary to set out the reasoning in the two cases in a little more detail.

The first is Mostyn v The West Mostyn Coal and Iron Company (1876) 1 CPD 145. The claim was brought in the Common Pleas Division and arose out of a lease. The lessee pleaded by way of defence that the lessor knew, and had concealed from the lessee, that he had no title to part of the land demised, and he counterclaimed to have the lease set aside. One of the issues was whether it was necessary to transfer the action to the Chancery Division. The Court held that it was not. Brett J said, at p. 150:

“If a defendant in an action in this Division sets up facts in his answer which in the Chancery Division would entitle him to have an instrument reformed or set aside, though this Division cannot reform or set it aside with regard to its effect in future, it may, for the purpose of determining the action, treat it as set aside.”

Archibald and Lindley JJ delivered concurring judgments. In my view the headnote accurately records the effect of the decision as a matter of principle:

“Where the defendant in an action in one of the Divisions of the High Court of Justice other than the Chancery Division relies on an equity to have a deed set aside as part of his defence, the Division in which the action is may give effect to the equity so far as these incidental to the purposes of the defence.”

The principle so stated is in accordance with the fusion of law and equity affected by the 1873 Act.

In the second case, Breslauer v Barwick (1876) 36 LT 52, that principle was applied in the context of a claim for rectification. The claim was brought in the Common Pleas Division by an individual claiming as charterer of a steamship. The defendant owner pleaded that it was not the plaintiff that was named as the charterer in the charterparty but one of his companies. The plaintiff in his reply pleaded that the company had been named by mistake and contrary to the intention of both parties. In his demurrer the defendant contended that the plaintiff could not proceed unless and until the charterparty was “reformed” (see n. 2 above), which required the case to be transferred to the Chancery Division. The Court rejected that argument. Brett J, with whom Grove J agreed, said:

“It is further said that the reply ought to ask that the charter-party be reformed, and that for this purpose the case be transferred to the Chancery Division; but the decision in Mostyn v The West Mostyn Coal and Iron Company shows that in such a case as this it is not necessary to go through the manual labour of reforming the agreement, but that if such facts are shown as would cause the Chancery Division to reform it, we may treat it as reformed, and give judgment accordingly.”

In other words, if the charterparty was rectifiable, applying the applicable principles, it would be treated for the purpose of the plaintiff’s claim as if rectified.

Those decisions are not binding on us; but it was not contended that they were wrongly decided, and the reasoning seems to me convincing. Although the principle which they state – in short, that rectifiability can be relied on as a defence even where the court has no power to rectify – was identified in the context of the jurisdiction of different divisions of the High Court I can see no reason why it should not apply generally, including where the jurisdictions in question are those of the High Court and the Employment Tribunal. It does not depend on the niceties of the drafting of the 1873 Act but on the distinction between raising an issue as a claim and raising it by way of defence: that distinction is equally valid in both contexts. It achieves a just outcome because it avoids what Brett J calls “the manual labour” (with the consequent expense and delay) of having to go to another court to seek rectification. An ET is as well able to decide a case of rectifiability for mistake as any of the other contractual issues which fall within its jurisdiction; indeed as a specialist tribunal it is likely to start with a better understanding of the context than a judge in the County Court or High Court.

I would add that, although the particular issue is different, such an outcome would tend in the same direction as the decision of this Court in Agarwal that an ET had jurisdiction to interpret the terms of a contract for the purpose of deciding a claim under Part II of the 1996 Act. At para. 27 of my judgment in that case I summarised my reasons for so holding as follows (so far as material):

“(1)

Delaney v Staples … is binding authority that an ET has jurisdiction to resolve any issue necessary to determine whether a sum claimed under Part II is properly payable, including an issue as to the meaning of the contract of employment.

(2)

...

(3)

There is no good – or even, frankly, comprehensible – policy reason for carving out from the jurisdiction of the ET one particular kind of dispute necessary in order to resolve a deduction of wages claim. On the contrary, to do so would be incoherent and would lead to highly unsatisfactory procedural demarcation disputes. ETs are well capable of construing the terms of employment contracts governing remuneration and have to do so in many other contexts.”

(The reference to Delaney v Staples is to the judgment of Nicholls LJ from which I have quoted at paras. 14 and 17 above.)

I accordingly believe that as a matter of law a case that the Letter Agreement was rectifiable for mistake could have been raised in Anderson by way of defence. To adapt Brett J’s language in Breslauer, if such facts were shown as would cause the High Court to make an order for rectification of the Letter Agreement the ET could have treated it as rectified and dismissed the complaint on that basis.

Although Males LJ agrees that Nexus could have advanced the mistake case in the ET, he does not base that conclusion on the principle exemplified by the Mostyn and Breslauer decisions, which he says has no role to play in the context of Part II of the 1996 Act: see para. 131 below. I respectfully disagree. The question whether an amount of wages is “properly payable” falls to be decided by reference to the general law, of which that principle is part. But I am not sure how substantial the difference between us really is. I do not disagree with anything he says in para. 132 of his judgment, and I agree with his proposition that a deduction may be lawful by reference to a contract which “needs to be rectified”. I accept that it may be possible to reach that conclusion by a purposive construction of the statute, but I think that the principle exemplified by the Mostyn and Breslauer decisions provides a firmer doctrinal basis for it.

It is fair to say that the Mostyn and Breslauer decisions were not referred to by either party below and were only drawn to the parties’ attention by the Court in the course of the hearing before us. The possibility of the ET determining a defence of rectifiability for mistake is not referred to in any employment law textbook of which I am aware; and in the only EAT decision to which we were referred in which it was argued that the contract of employment did not reflect the parties’ common intention it was assumed that the issue would have to be decided by a claim for rectification in the County Court – see para. 24 of the judgment of Burton Jin Mango Hair v Dos Santos (2002) EAT/129/02. It may therefore seem rather hard to say that the mistake could with reasonable diligence have been advanced. But even if it was not reasonable to expect anyone to spot those particular authorities, or to apply the proposition in Snell (see para. 53 above) to a situation like the present, I believe that it would not have required more than reasonable diligence to advance the mistake case on the basis preferred by Males LJ. And in fact in the Bolam proceedings Nexus did plead precisely such a case. Paras. 13.3 and 13.4 of its Grounds of Resistance raise by way of defence both a case of common mistake and, in the alternative, a case of unilateral mistake. As regards common mistake it is averred that the Letter Agreement was

“void in so far as it is said to operate to increase shift pay in the manner alleged by the Claimants”.

As regards unilateral mistake it is averred that the Agreement

“is either void ab initio by reason of unilateral mistake on the part of the Respondent or, alternatively, it is binding only to the extent intended by the Respondent or it is susceptible to rectification or rescission and it is unconscionable for the Claimants to assert to the contrary”.

The references to rescission and to the Letter Agreement being “void” do not correspond to my analysis above, but the reference to the Agreement being “susceptible to rectification” is consistent with both my approach and that of Males LJ. (Footnote: ⁶)

As to how the Unions put their case before the Judge, Lord Hendy, as I understand it, relied primarily on the terms of the Bolam pleading: his argument was that since Nexus pleaded a mistake case in Bolam it could equally have done so in Anderson. In his skeleton argument before us Lord Hendy continued to advance at least part of that argument, contending that it would have been open to Nexus to argue in Anderson that its mistake rendered the Letter Agreement “unenforceable” or “void”, with the result that the enhanced payments of shift allowance rate were not properly payable. As already noted, that terminology does not seem to me quite correct; but after considering Mostyn and Breslauer Lord Hendy modified his analysis to put the emphasis on rectifiability.

The Judge addressed the question whether Nexus could have advanced the mistake case in the Anderson proceedings at paras. 31-32 of his judgment, as follows:

“31.

I reject the defendants’ submission. The mistake, if there was one, arose in the context of the negotiation of the collective agreement. Rectification of the Anderson claimants’ individual employment contracts (or the collective agreement) was not possible in the Anderson proceedings. It is difficult to see how the issue of mistake, whether common or unilateral, could be satisfactorily resolved given that the Anderson claimants were not parties to the collective agreement and the defendants were not parties to the Anderson claimants’ employment contracts. Factors which might come into play in the exercise of the court's discretion whether or not to grant rectification would not have been relevant in the Anderson proceedings.

32.

For those reasons, notwithstanding the claimant’s grounds of resistance in the Bolam proceedings, I do not consider that it was possible with reasonable diligence for the claimant to have raised the mistake arguments in the Anderson proceedings. Even if [it] had been possible, in my judgment, those arguments were not ones which in all the circumstances the claimant should have raised. This outcome also works justice since in my judgment it would not be just to deprive the claimant of the opportunity to advance the rectification claim against the defendants, whatever may be the claim’s merits, on the basis of the outcome of the Anderson proceedings. Were the claimant estopped from pursuing a well-founded claim for rectification, it would mean that the outcome of the Anderson proceedings would have significant financial consequences in respect of every employee of the claimant in whose employment contract the Letter Agreement was incorporated.”

I have quoted the whole of para. 32 for convenience, but in fact only the first sentence is concerned with the “could” question: I deal with the remainder, which addresses the “should” question, at paras. 67-71 below.

The essential reason for the Judge’s decision that it was “not possible” to raise the mistake case in the Anderson proceedings was that the parties to those proceedings, i.e. the individual employees, were not parties to the Letter Agreement. I respectfully disagree with that approach, for the reasons which appear under head (A) above: it is the individual contracts, not the Letter Agreement, that are the proper target of any decision about rectifiability. Once that is appreciated, there would, for the reasons which I have given, have been no obstacle to the ET deciding that issue; and indeed the Judge does not rely on any problem of jurisdiction of that kind. He says that it would not be possible “satisfactorily [to] resolve” the issue of mistake where the individual employees themselves were not parties to the Letter Agreement, but I do not see why. Of course the relevant evidence would be that of the Union negotiators, but there would have been no difficulty about adducing that evidence in the ET: the RMT were in practice running the proceedings. Any discretionary factors of the kind to which the Judge refers in the final sentence of paragraph 31might also relate more to the conduct of the Union negotiators than to that of the employees; but again there is no reason why such factors could not have been considered by the ET.

Mr Reade’s primary submission before us was that the mistake case could not have been advanced in the ET because it had no jurisdiction to order rectification. He was constrained to acknowledge that a version of the mistake case had been advanced in Bolam, but he said that the case as there pleaded did not work as a matter of law. As regards the analysis based on Mostyn and Breslauer, he emphasised that, even if (contrary to his primary position) it applied in the ET, a defence of “rectifiability” was limited, as Snell puts it, to “past transactions”.

I would reject Mr Reade’s primary submission for the reasons which I have already given. The ET did have jurisdiction to determine the mistake case as part of a defence to the Anderson claim, and although the details of the pleading in Bolam may be questionable its overall approach was correct. The fact that rectifiability is limited to past transactions is not a reason why it could not have been raised in the Anderson proceedings. If it had succeeded it would have defeated the actual complaints before the ET, i.e. the complaints about the pre-complaint deductions. Further, though this is not the essential point, although the tribunal could not formally have rectified the Letter Agreement its decision would in practice (subject to any appeal) have made it impossible for the claimants to bring any claims as regards future payments of shift allowance at the unenhanced rate.

Mr Reade submitted by way of fallback that even if the ET had jurisdiction to determine the mistake case Nexus could not with reasonable diligence have advanced it. I have already addressed one aspect of that point: see para. 60 above. But in fact the basis for the submission in his skeleton argument was the proposition that the ET could not order rectification and/or that the parties to the Anderson proceedings were not the same as the parties to the Letter Agreement. I am not sure that those points go to the question of reasonable diligence, but in any event I do not accept them, for the reasons already given.

Such evidence as there is about why the mistake case was not raised in the Anderson proceedings appears at para. 61 of the witness statement of Mr David Bartlett, the head of Nexus’s Business Change and Technology Department, who says that “following [the decision of this Court in Anderson] … senior counsel raised the possibility of the claimant pursuing a claim for rectification against the unions on grounds of common or unilateral mistake”. That must, I think, mean that that was the first time that the possibility of advancing the mistake case was considered: the common sense inference must be that the possibility of advancing the mistake case simply did not occur to Nexus or those advising them in the course of the ET proceedings.

In my view the mistake case should have been advanced in Anderson. In the absence of some special reason it is plainly incumbent on a party to advance all available defences. As we have seen, no such reason is suggested by the evidence. This is not a case where there had been some subsequent development – either factual or legal – which justified Nexus in not doing so. In ordinary litigation it is common, where a case turns on a disputed issue of construction of a written instrument, for a party to aver that if their construction is found not to correspond to the meaning of the document it falls to be rectified for mistake (assuming, of course, that they have an evidential basis for the averment). It may be less usual for such a situation to arise in ET proceedings, but it can happen (as the case of Mango Hair illustrates) and there is no reason for any different approach in that context. There is clearly an overlap with the issue of “reasonable diligence”: to the extent that there is, what I say above applies equally.

The Judge addressed the “should” question in para. 32 of his judgment: see para. 62 above. As I read it, he regarded his conclusion in the second sentence as justified by the same reasons as his conclusion on the “could” issue, with which I would respectfully disagree for the reasons already given.

The second and third sentences of para. 32 deal with an argument to the effect that the outcome is “also” just in a broader sense. That is apparently intended only as a supportive point and it will be convenient to deal with it later: see paras. 83-84 below.

Newey LJ reaches a different conclusion on the “should” issue: see para. 117 below. I note his concern that to raise rectifiability by way of defence rather than by way of counterclaim for rectification is unsatisfactory, and I would not disagree where the issue arises in a court which has power to order rectification; but for the reasons already given I believe that in a case like the present it is unobjectionable. Males LJ does not believe that we are in a position to decide the “should” issue ourselves: see para. 137. Fortunately, however, as will appear, these difference do not affect the outcome of the appeal.

Finally on this aspect, I should refer to a decision relied on by Mr Reade before us, though not before the Judge, which is arguably relevant to both the “could” and the “should” questions. Crane v Hegeman-HarrisCo Inc [1939] 4 All ER 68 concerned a dispute about remuneration between an architect and a building contractor. The dispute was referred to arbitration. The arbitration proceeded on the basis that the sole question referred related to the construction of the remuneration clause in the relevant agreement. The arbitrator determined that question in favour of the architect. When he brought proceedings to enforce the award the contractors served a defence claiming that the term in question did not express the true agreement between the parties and counterclaimed for rectification. The architect argued (among other things) that they were estopped from doing so. There is thus at first sight some similarity with the issue before us.

Lord Greene MR, who delivered the leading judgment, upheld the decision of Simonds J ( [1971] 1 WLR 1390 (note) (Footnote: ⁷)) that the terms of the submission to arbitration were limited to disputes arising out of the document purporting to record the remuneration agreement and did not cover any issue as to whether that document reflected the true agreement between the parties. He continued, at p. 72 E-G:

“The only other point is … a suggestion that there was some sort of estoppel or quiescence by the respondents which deprives them of the right at this stage to raise the point. [Counsel for the claimant] contended that a party in circumstances such as these was not entitled to take two bites at a cherry. I think it is sufficient to say that nobody in the wildest flight of metaphor would say that a person was bound to take one bite at two cherries. The issue of rectification is a totally different issue. The only issue before the arbitrator related to the actual document, and there was nothing in the world to prevent the respondents from proceeding with that issue which was submitted until its conclusion, and then to raise a different issue altogether which was entirely outside the jurisdiction of the arbitrator.”

The way that the estoppel point is there formulated appears to be rather different from how it was understood by Simonds J, who treated it as being based on estoppel by representation and could find no relevant representation; but as stated by Lord Greene it bears some similarity to Nexus’s argument before us. However, the crucial distinction is that his reasoning is squarely based on the fact that the arbitrator had no jurisdiction to determine the rectification issue (because of the limited scope of the submission), whereas in our case the ET did have jurisdiction to determine rectifiability. That being so, I do not believe that Crane is relevant authority. I would add that the case pre-dates the clarification of the law of what we now label “issue estoppel” (a term which did not enter the English jurisprudence until Thoday v Thoday in 1964): the architect’s case might have been put rather differently in the light of Arnold and Zodiac, though the limitation on the arbitrator’s jurisdiction might still have posed an insuperable problem.

If in Anderson the ET had decided not only the issue of principle but also the amount of the pre-complaint deductions in each case an order for rectification of the Letter Agreement could not assist Nexus as regards those amounts: there would be a final order which could not be challenged except by appeal (or by seeking to have it set aside for fraud). However, as we have seen, no such determination has occurred because the ET deferred the quantification of the underpayments in each case (described as “remedy”). That being so, Mr Reade confirmed to us that it is Nexus’s intention, if it succeeds in its rectification claim, to argue at the “remedy” stage, unattractive as it might appear, that it follows from that fact that no amount is payable notwithstanding the ET’s previous decision. The question is whether such an argument would be precluded by res judicata or would constitute an abuse of process.

Mr Reade was right to acknowledge that Nexus’s stance is unattractive. In Fidelitas Shipping Co Ltd v V/O Exportchleb [1966] 1 QB 630 Diplock LJ said, at p. 642, that it was “too clear to need citation of authority” that where an issue raised by a claim is dealt with as a preliminary issue

“… the parties to the suit are bound by the determination of the issue. They cannot subsequently in the same suit advance argument or adduce further evidence directed to showing that the issue was wrongly determined. Their only remedy is by way of appeal from the interlocutory judgment … This is but an example of a specific application of the general rule of public policy, nemo debet bis vexari pro una et eadem causa. The determination of the issue between the parties gives rise to what I ventured to call in Thoday v Thoday [1964] P 181, 198, an ‘issue estoppel’.”

On the face of it, a case where a court has made a finding of “liability” but deferred a decision on quantification is the paradigm of the kind of situation that Diplock LJ was referring to. However, I think it is necessary to analyse the position more fully. I will address in turn cause of action estoppel, issue estoppel and abuse of process.

I start with two potential objections to the application of cause of action estoppel as regards the pre-complaint deductions.

First, Mr Reade submitted that the cause of action determined by the ET does not correspond to the subject-matter of the claim in this action. He submitted that the ET was concerned with a statutory claim that Nexus had made unlawful deductions from the claimants’ wages within the meaning of Part II of the 1996 Act, whereas the issue in this action concerns their contractual entitlements, and that accordingly cause of action estoppel does not operate. I do not accept that submission. The distinction between the subject-matters of the two claims is in my judgment nominal rather than substantial. The substance of the ET’s finding in Anderson is that the sums in question are due under the employees’ contracts; all that the statute does is to afford an alternative mechanism for vindicating the claim. The substance of the decision sought by Nexus in these proceedings (as regards the pre-complaint deductions) is that the selfsame sums are not due under the contract. Thus the ultimate claim is the same in both sets of proceedings.

Secondly, I have been concerned whether the fact that the ET has made no final determination of the claim of any individual means that no cause of action estoppel can arise: the cause of action is the non-payment of a specific sum, and no such sum has been quantified in any of the claimants’ cases. If that is so, then we are concerned with issue estoppel rather than cause of action estoppel: which characterisation is correct might in principle make a difference because, as we have seen, issue estoppel is potentially more flexible in its application than cause of action estoppel. In my view, however, the considerations of policy which underlie the doctrine of cause of action estoppel justify its application in a case of the present kind, where a court or tribunal decides that a claimant has a cause of action and simply defers quantification: what matters is that there has been a definitive decision on the question of liability. I note that in the passage from his judgment in Fidelitas quoted above Diplock LJ referred to the operative principle as issue estoppel; but I think a case of the present kind is to be distinguished from a conventional “preliminary issue”, where the determination of the issue does not determine liability. It must also be borne in mind that Diplock LJ adopted the (avowedly novel) terminology of issue estoppel at a time when the taxonomy of this subject had not been fully developed as it has since been in Arnold and Zodiac.

Once that point is reached, since I have already held that Nexus could and should have raised the mistake case in Anderson, it would follow that it is precluded by cause of action estoppel from contending at the remedy stage that shift allowance was not payable at the enhanced rate.

That conclusion means that it is strictly unnecessary to consider the question of issue estoppel. I will, however, do so in case I am wrong on either of the two questions considered at paras. 78 and 79 above.

The primary requirements for an issue estoppel are plainly present. The ET found that the employees were contractually entitled to enhanced shift payments, whereas the purpose of the rectification claim is to enable Nexus to show, inconsistently with that finding, that they were not so entitled. (If it is necessary to put the point in Lord Sumption’s terms, the fact that the terms of the Letter Agreement represented both parties’ intentions was “essential to the existence of the cause of action” in Anderson.) I have already held that the mistake case could and should have been advanced in the Anderson proceedings. The only question is thus whether there are special circumstances which mean that the application of the bar would cause injustice.

As to that, I should start by making the point that, while in one sense it might be said to be unjust that Nexus should be precluded from advancing what may be a well-founded defence, that possibility is inherent in any case of the present kind and cannot by itself constitute a relevant injustice. It will only have arisen because Nexus failed to raise in the Anderson proceedings a point which could and should have been raised; and doing justice to both parties requires proper weight to be given to the important interest in finality. Accordingly what Nexus needs to show is some special circumstance which would render it unjust for it to be debarred from advancing the mistake case notwithstanding the importance of finality.

In the present case the interest in finality is particularly strong. Not only did Nexus not advance the mistake case in the ET, it failed to do so either in the EAT or before this Court. The only reason why the point remains even potentially open to it is that as a matter of case management final quantification was deferred pending the outcome of those appeals. The present proceedings were commenced only after all appeals had been exhausted (save for the application for permission to appeal to the Supreme Court), almost five years after the start of the ET proceedings (and, be it noted, almost eighteen months after Addleshaw Goddard’s letter before action). In those circumstances it is very hard to see how it could be just to allow the mistake case to be advanced now.

The Judge addressed this issue at para. 35 of his judgment, as follows:

“For the reasons stated above in the context of cause of action estoppel, the claimant is not issue estopped from pursuing the rectification claim here. The mistake issue was not ‘necessarily common’ to the Anderson proceedings and the present proceedings; nor could it with reasonable diligence or should it in all the circumstances have been raised in the Anderson proceedings. In view of the significant financial consequences referred to above, I would, had it been necessary to do so, therefore have held that special circumstances exist in which it would not be just to deprive the claimant of the opportunity to advance the rectification claim against the defendants.”

It will be seen that the Judge there largely adopts his previous reasoning in relation to cause of action estoppel. I have already addressed that reasoning, and it will be apparent why I respectfully disagree with him. But I should say something about his reference to “significant financial consequences”, amplified in para. 32 as “significant financial consequences in respect of every employee of the claimant in whose employment contract the Letter Agreement was incorporated”. It is clear that what he had in mind was that Nexus would be exposed to a liability to pay shift allowance at an enhanced rate to all the relevant employees for an indefinite future period. That, however, is not the liability with which we are concerned at this stage of the argument, which is limited to liability for the pre-complaint deductions claimed in the Anderson proceedings: I return later to the question of liability for the post-complaint deductions and to other employees.

It is clearly not enough to say that it is unjust that Nexus should have to pay a large sum for which (ex hypothesi) it might not be liable; that would undermine the importance of finality which issue estoppel is intended to protect. It may be – I express no concluded view – that a different conclusion would be justified if the sum involved were quite exceptionally high, or such as to imperil the continuation of Nexus’s business. But there is no evidence to that effect. Indeed the statement of facts says nothing about the financial consequences of the ruling. In his judgment in Anderson Employment Judge Hunter refers to the amount in issue to date as being of the order of £500,000, representing an average increase in pay for the relevant employees of 5%. The parties were asked by the Judge whether that was an agreed figure. Lord Hendy did not accept it, but I am prepared to proceed on the basis that the amount is of at least approximately that order: that would not be very surprising given that 70 employees were claiming to have suffered regular shortfalls over a two-year period. I do not believe that the fact that Nexus would have to pay such a figure can possibly outweigh the interests of finality as regards the pre-complaint deductions.

Accordingly, even if Nexus were not precluded by cause of action estoppel from relying on a successful rectification claim to defeat the claims of the Anderson claimants in respect of the pre-complaint deductions, I have no doubt that they would be precluded by issue estoppel.

Finally, even if, contrary to my view, it cannot be said that Nexus should in all the circumstances have advanced the mistake case for determination by the ET, so that res judicata is not available, I believe that any attempt to rely on that case in respect of the pre-complaint deductions would be an abuse of process. The only responsible course for Nexus was to raise the mistake issue at an early stage in the Anderson proceedings, albeit on the basis that it represented a fallback, so that a decision could be made as to the appropriate case management options. I strongly suspect that the ET would have stayed the proceedings pending the outcome of proceedings in the High Court which could have decided both the construction issue and (if necessary) the claim for rectification: that was what the EAT decided should be done in Mango Hair (Footnote: ⁸).

In summary, any attempt by Nexus at the remedy stage to defeat an order for payment of the pre-complaint deductions on the basis of the mistake case would be precluded by cause of action estoppel or issue estoppel; and even if I were wrong about that it would fall to be dismissed as an abuse of process.

Although neither Males LJ nor and Newey LJ agrees with all aspects of my reasoning on res judicata both are agreed that it would be abusive for Nexus now to rely on rectification (if proved) as a defence to the Anderson claimants’ claims for the pre-complaint deductions. Thus, by whichever route, Nexus cannot rely on rectification as a defence to those claims.

The Anderson decision cannot give rise to a cause of action estoppel as regards the post-complaint deductions because the ET was formally only concerned with underpayments which had occurred prior to the presentation of the complaint. Accordingly, if Nexus in due course sought to defend a claim for the post-complaint deductions by the Anderson claimants on the basis that the Letter Agreement had been rectified in the present proceedings, any answer based on res judicata would have to take the form of issue estoppel.

As in the case of the pre-complaint deductions, the primary requirements of issue estoppel are plainly present (see para. 82 above), and I have already concluded that the mistake case could and should have been advanced in the Anderson proceedings. Again, therefore, the only question is whether there are special circumstances which mean that it would be unjust for Nexus to be prevented from seeking to establish its mistake case despite its failure to raise it in Anderson.

I can see the case that there was for this purpose no relevant distinction between the claims for pre-complaint and post-complaint deductions. Notwithstanding the formal position, both parties must have understood that the outcome of the Anderson proceedings would in practice determine the issue of the claimants’ entitlement to enhanced shift allowance on an ongoing basis (see para. 18 above); and there is nothing unjust in Nexus being precluded from re-fighting that issue as regards further deductions made on the identical basis.

That is a powerful point, but I do not think that it is necessarily determinative. If it is not rectified, the Letter Agreement will remain part of the contracts of employment of the relevant employees – at least unless and until re-negotiated, which it is hard to see that the employees or the Unions would have any incentive to do. It is true that the cohort of Anderson claimants will gradually diminish as they leave the workforce (Footnote: ⁹), but it may in this context be relevant to look beyond that cohort to all employees in the relevant grades, past and future (since it is not clear whether it is realistic for Nexus not to apply the Letter Agreement to incoming employees). Accordingly if it is precluded from relying on the mistake case, Nexus will have to pay the enhanced rate of shift allowance to a large number of employees for an indefinite and potentially very long period, in circumstances where (as we have to assume at this stage) there is an arguable case that that is not what either they or the Unions intended: this is the point that I understand the Judge to have been making in the second part of para. 32 of his judgment. In my view it is arguable that such an outcome would, if the mistake case were proved, give disproportionate weight to the interests of finality, important though they are. What is required is a holistic assessment of where the balance of justice lies between those interests and the impact on Nexus of not being able to advance its case. I do not believe that the balance necessarily falls the same way as regards the pre-complaint and the post-complaint deductions.

It is important to make clear that I do not believe that the Court has to choose between the two extremes of, on the one hand, preventing Nexus from relying on the mistake case for ever and, on the other, allowing it to do so in respect of the entirety of the post-complaint deductions. On the contrary, it would be open to it to find that it was just for the issue estoppel to operate as regards claims for deductions up to, but only up to, some specified date later than the commencement of the Anderson proceedings. I express no view about what such a date might be: possible candidates are the promulgation of the ET decision, the sending of the letter before action and the commencement of the present proceedings, but there may be others. An alternative, and perhaps more appropriate, way of achieving the same result would be by making any order for rectification on terms that prevented Nexus from relying on it as an answer to deductions made before the specified date. Terms of that kind are entirely appropriate in the context of an equitable remedy: see, for example, para. 40 ofLord Neuberger’s judgment in Marley v Rawlings [2014] UKSC 2,  [2015] 1 AC 129, where he said:

“If it is a question of interpretation, then the document in question has, and has always had, the meaning and effect as determined by the court, and that is the end of the matter. On the other hand, if it is a question of rectification, then the document, as rectified, has a different meaning from that which it appears to have on its face, and the court would have jurisdiction to refuse rectification or to grant it on terms (e.g. if there had been delay, change of position, or third party reliance).”

I do not believe that this Court is in a position to perform the necessary holistic exercise. The submissions before us did not approach the issue in that way, and the statement of facts does not necessarily provide an adequate basis for it. Even if we had more information than we do, I would have been inclined to regard the exercise as premature at this stage. I can understand why the Unions would like to have the res judicata issue decided by way of preliminary issue, in the hope that that would avoid the expense and delay involved in having a trial of the mistake case, and in many or most cases that would indeed be the appropriate course; but in the particular circumstances of the present case I think it is better for it to be determined (if it arises) by the Court hearing the rectification claim.

If, contrary to my view, there was no prima facie issue estoppel as regards the post-complaint deductions, my conclusion would be the same about whether this Court could decide whether Nexus’s reliance on the mistake case was an abuse of process. That is unsurprising, since, as Lord Sumption recognises, the same underlying principles are involved.

It follows from that approach that I believe that the Judge was wrong definitively to decide the issues of res judicata and abuse of process as regards the post-claim deductions.

I turn finally to an issue which on the basis of my conclusion on issue (A) does not arise. If, contrary to that view, Nexus were in fact entitled to proceed against the Unions alone, a defence of res judicata could still be raised if they are in law the privies of the Anderson claimants and thus entitled to rely on the decision in their case. We were addressed fully on that issue, and I think I should deal with it.

In my view the Unions would indeed be entitled to rely on the Anderson decision on the basis that there was privity of interest between them and the individual employees who claimed in those proceedings. The Anderson proceedings were concerned with the meaning and effect of a collective agreement which the Unions had negotiated on behalf of the bargaining unit to which the claimants belonged. In the negotiation the Unions were not acting in their own interests but solely in those of the employees in question: they were their representatives. The effect of the collective agreement was (and was only) to create for the employees the legal rights which were the subject of the Anderson proceedings. It is true that in law the Unions were not acting as the employees’ agents, but, as appears from element (c) in Floyd LJ’s summary in Resolution, we are concerned with substantive justice; and the relationship is in my view in this context substantively identical. Adapting Sir Robert Megarry’s formulation in Gleeson, there is a sufficient degree of identification between the Unions and the employees to make it just to hold that res judicata should be available as between Anderson and the current proceedings.

The Judge reached the contrary conclusion. His reasons appear at para. 43 of his judgment, as follows:

“Adopting the approach taken by Floyd LJ [in Resolution], I accept that the defendants had an interest in the subject matter of the Anderson proceedings. However, I do not accept that the defendants can be said to be, in reality, the party to the Anderson proceedings. The facts that the Anderson claimants were members of the RMT, that their claim was being supported and funded by the RMT and that the Letter Agreement was in issue in the Anderson proceedings and is in issue in the present proceedings are, in my judgment, insufficient for that purpose. The RMT could not have been a party to the Anderson proceedings and could not have brought the claim which the Anderson claimants brought. Whether or not the defendants could be said to be, in reality, the party to the Anderson proceedings, I do not consider that it would be just that the claimant should be bound by the outcome of the Anderson proceedings, for the reasons already stated. In my judgment, the fact relied on by the defendants that, from a practical viewpoint, the construction of the Letter Agreement arrived at in the Anderson proceedings would affect the collective negotiations going forward is a factor in favour of the claimant and not that of the defendants in determining whether it would be just that the claimant should be estopped from pursuing the rectification claim.”

With respect, I do not accept that reasoning. For the reasons which I have given, the fact that the Anderson proceedings and the present proceedings are concerned with what contractual rights were created for the employees by the Letter Agreement which the Unions negotiated is sufficient to establish a privity of interest. (To the extent that the Unions placed primary weight on the facts that the Anderson claimants were members of the RMT and that their claim was being supported and funded by it, I agree that those facts are not relevant as such, though they may illustrate the nature of the relationship.) Although the Judge is right to say that the RMT (strictly the Unions) could not in law have been a party to the Anderson proceedings, I do not believe that element (b) in Floyd LJ’s summary in Resolution means that it is a requirement that that should be the case: the test is a broader one of whether the Unions’ interest in those proceedings is sufficient to make it just that they should be able to rely on the ET’s decision in order to raise a defence of res judicata in the present proceedings. The Judge does in fact address that question in the second part of the paragraph. His reference to “the reasons already stated” is to the substantive question of whether res judicata is established: I do not believe that that bears on the privity issue.

Nexus is not entitled to rely on the mistake case, if proved, as an answer to a claim for the pre-complaint deductions. The question whether, and if so to what extent, it is entitled to do so as regards the post-claim deductions should be decided (if necessary) by the Court hearing the rectification claim.

OTHER EMPLOYEES

We are here concerned with employees in grades 1-3 who were not claimants in the Anderson proceedings. Some, as we have seen, have since brought their own claims – see para. 10 above – and there are others who may yet do so. Nexus wishes in any such claims to advance the mistake case by way of defence, as it already has in Bolam. The question is whether it is precluded from doing so by res judicata or, even if it is not, whether to do so would be an abuse of process.

The employees in question would not in those circumstances be entitled to rely on res judicata because they were neither parties to the Anderson proceedings nor the privies of the claimants in those proceedings. In his reply Lord Hendy submitted, for the first time, that all employees in grades 1-3 were in privity with one another because they were all negotiated for by the Unions. I do not accept this. The employees in the bargaining unit do not have a relationship of the kind required by the authorities. They are simply individuals who enjoy similar contractual rights against a third party. The outcome of the Anderson proceedings may for that reason be “of concern” to the other employees because it suggests that similar claims by them will succeed, but, as Sir Robert Megarry says in the first paragraph quoted from Gleeson at para. 43 above, that is not enough to constitute a privity of interest.

Accordingly the employees would have to rely on abuse of process. As to that, it is no doubt arguable that it must have been understood by both the Unions and Nexus that the Anderson proceedings would determine the entitlement of all relevant employees, both current and future, to enhanced shift allowance; and that it would accordingly be an abuse of process for Nexus, having lost in those proceedings, to seek to defend claims by other employees on a basis not advanced in Anderson. Lord Hendy referred us to Ashmore v British Coal Corporation [1990] 2 QB 338, in which this Court upheld the striking-out of a claim by an employee where a substantially identical claim had previously been dismissed in proceedings involving other employees. The circumstances are not identical to those of the present case because the claim in question was one of a large group which had been case managed together and in which the cases which were the subject of the earlier decision were selected as lead cases; but although that feature is not present here Lord Hendy submitted that that distinction was not fundamental.

Mr Reade argued that it would be wrong to equate the position of the other employees with that of the Anderson claimants, precisely because they were not parties to the Anderson proceedings. He reminded us of Thomas LJ’s statement in Aldi, quoted at para. 48 above, to the effect that difference of parties was a “powerful factor” weighing against a finding of abuse in cases of this kind. But that may not be a strong point in the particular circumstances of the present case, where the Letter Agreement was negotiated collectively for the benefit of a defined bargaining unit: in those circumstances it might be thought to be natural for all parties to expect that a decision by a tribunal as to its meaning and effect would apply to everyone in that unit.

I do not believe that we should attempt to resolve those issues. Whether or not the other employees are in a relevantly worse position than the Anderson claimants as regards the post-complaint deductions, they clearly cannot be in a better position. In their case we have even less information than in the case of the Anderson claimants on which to decide how the relevant balance of justice should be struck. In particular, there may be a variety of reasons why they did not join in the Anderson proceedings which may be relevant to the question whether or to what extent they should be entitled to prevent Nexus relying on any rectification of the Letter Agreement. That question is, again, best decided (if it arises) by the Court deciding the rectification claim.

I would allow the Unions’ appeal against the Judge’s dismissal of the strike-out application on ground (iii), and I would accordingly dismiss the action. It is for Nexus to decide whether to bring fresh proceedings against individual employees.

If Nexus does bring fresh proceedings and succeeds in proving the mistake case, it follows from paras. 90-91 above that it will nevertheless not be entitled to rely on rectification as defeating the claims of the Anderson claimants as regards the pre-complaint deductions. It follows that there is no reason why the quantification of the individual claims by the ET, if they are not agreed, should not proceed (though it is not within the jurisdiction of this Court to make an order to that effect). As regards the other claims, if the Court finds the mistake case proved it will have to decide at that stage whether and to what extent Nexus is precluded from relying on rectification as an answer to those claims. The sooner there is a trial of the mistake case the better.