Rana v London Borough of Ealing & Anor

These two appeals – to which I will refer as Rana and Bonnie – are against separate decisions of HH Judge Eady QC in the Employment Appeal Tribunal refusing to extend time for lodging an appeal against a decision of the Employment Tribunal. They were heard together because they raise a common point of law about the calculation of the time limits for such an appeal. Under the relevant rule time starts to run where the ET’s judgment and/or the written reasons are “sent to the parties”. In both the cases before us they were erroneously sent to a former representative of the party wishing to appeal: in a nutshell, the question is whether they were nevertheless “sent to the parties” for the purpose of the rule, and, if they were, what approach should be taken to extending time. We were able to reach a decision before we were in a position to give our reasons, and in order to avoid further delay judgment was handed down on 6 September allowing both appeals with reasons to follow. These are my reasons for that decision.

Despite that common issue it is necessary to consider the two appeals separately. In Bonnie both parties were represented (the Appellant by Mr Saul Margo and the Respondent by Mr Robert Moretto) whereas in Rana the Respondents made the choice not to instruct counsel but to rely on their solicitors’ skeleton arguments, with the result that only the Appellant was represented (by Mr William Young). In those circumstances by agreement between counsel Mr Margo made his submissions before Mr Young, and the argument in Bonnie was inevitably more extensive. That being so, I will take Bonnie first, although it is in fact the later in point of time; but I will in dealing with the common issue refer so far as necessary to the submissions of Mr Young and the Respondents in Rana.

The quality of the written and oral submissions by all counsel was high, but Mr Moretto will understand if I express the Court’s particular gratitude to Mr Margo and Mr Young, who both appeared pro bono.

The provision governing the time limit for appeals to the EAT is rule 3 (3) of the Employment Appeal Tribunal Rules 1993 (as amended). The rule is rather elaborate, because it deals separately with a variety of different ways in which the ET’s judgment or the reasons for it may be promulgated. But the relevant part for our purposes reads:

The time limit in rule 3 (3) can be extended under the general discretion available to the EAT under rule 37, exercisable in the first instance by the Registrar but on appeal by a Judge. The principles governing the exercise of that discretion by the EAT, and its review on appeal, have been considered by this Court on several occasions, most recently in Green v Mears Ltd [2018] EWCA Civ 731, in which I gave the leading judgment. It is not necessary to repeat the exercise which I carried out there. In short, the principles first enunciated by Mummery J in Abdelghafar v United Arab Emirates [1994] ICR 6, as expounded in Jurkowska v Hlmad Ltd [2008] EWCA Civ 31, [2008] ICR 841, continue to apply. (Footnote: ¹) Very broadly, if there is no good explanation for the failure to meet the deadline for appealing, it is exceptional for the discretion to extend time to be exercised; and a strict view is taken of what constitutes a good explanation or exceptional circumstances.

Two other provisions, or groups of provisions, are potentially relevant to the issue before us.

First, on 29 July 2013 the President of the EAT issued a Practice Direction under powers contained in section 29A (1) (b) of the Employment Tribunals Act 1996. Paragraph 5.3 deals with the time limit applicable to appeals against a judgment. Most of it is immaterial for our purposes, but it concludes (so far as relevant):

Secondly, it is necessary to be aware of the provisions of the Employment Tribunal Rules of Procedure 2013 about the promulgation of judgments and written reasons. Rule 61 (2) provides that where a judgment is not given at a hearing it will be reserved “to be sent to the parties as soon as practicable in writing”. Rule 62 (2) reads (so far as material):

It is not necessary to say anything about the substance of the Appellant’s claims, which were of unfair dismissal and disability discrimination. In her ET1 she nominated a firm of solicitors as her representative and gave their address; as the form makes clear, any correspondence would thereafter be with them rather than with her, as long as they remained her representatives. She subsequently notified the tribunal that they were no longer acting for her and asked for all communications to be sent to her home address (which was the same as appeared in the ET1). At some later point a firm called Grand & Machyle took over as her representatives.

The hearing of the Appellant’s claims took place over two weeks in the London Central Employment Tribunal, starting on 19 January 2015. On 21 January Grand & Machyle wrote to the tribunal coming off the record, and thereafter the Appellant represented herself. They did not in that letter explicitly notify the tribunal of a different address at which it or the Respondent should communicate with the Appellant, as required by rule 86; but it has not been argued before us that their address remained valid for those purposes, and it appears to be common ground that the Appellant’s address thereafter for the purpose of the rule was either her postal address, as given in the original ET1, or her e-mail address (which she had given to the tribunal in the course of the hearing).

At the conclusion of the hearing the ET reserved its decision. On 28 April 2015 the judgment and written reasons were signed by Employment Judge Tayler. In a case where the decision is reserved these form a single document, and for convenience in this case (though not in Rana) I will refer to both as “the judgment”. The judgment ends with the standard endorsement (see para. 7 above) recording that it was sent to the parties on the same day that it was signed.

In fact, however, the Appellant’s copy of the judgment was sent not to her in person but to Grand & Machyle, who were no longer acting. The Respondent’s solicitors spotted straightaway what had happened and, sensibly and in accordance with the over-riding objective, e-mailed the tribunal, the Appellant herself and Grand & Machyle to point out the mistake: they gave the tribunal the Appellant’s e-mail address. Grand & Machyle do not appear to have responded. More remarkably, and most reprehensibly, the tribunal did not respond either, and the Appellant herself had to write on no fewer than five occasions, and make numerous telephone calls, before finally on 4 June receiving by e-mail (without any apology or explanation) a copy of the judgment, about five weeks later than she should have.

As soon as she received a copy of the judgment, the Appellant telephoned the EAT to ask how to appeal, and she was sent by e-mail the same day a copy of the Practice Direction and relevant Rules. On 17 June she made an application to the ET for a reconsideration, which was in due course refused.

On 15 July 2015 the Appellant filed her appeal at the EAT. She was notified that the appeal was 39 days out of time – as it was, if the 42 days ran from 28 April – and she was invited to apply for an extension. She did so. The application was dismissed by the Registrar on the papers. The Appellant appealed from that decision. At the hearing before Judge Eady she represented herself. The Respondent was represented by counsel (not Mr Moretto). It is not entirely clear whether the Appellant gave formal evidence on oath, but she had submitted evidence in the form of correspondence, with attached documents, and she made oral submissions.

Since an appeal against the order of the Registrar is by way of rehearing we are only concerned with the decision of the Judge. Judge Eady’s reasoning can be summarised as follows. She accepted that the Appellant had not received the judgment until 4 June 2015, and that it was not in any way her fault that she had not received it earlier. She found that while the member of the EAT staff to whom she spoke on 4 June may have told her that she had 42 days to appeal, starting from when the judgment was sent out, the Appellant had not explained to them the circumstances outlined above – i.e. that there had been a five-week delay in it being sent to her personally – and the person to whom she spoke could not reasonably have been understood to have advised that in those circumstances time only ran from 4 June. She continued, at para. 17 of her judgment:

The Appellant’s original grounds of appeal were drafted apparently without legal assistance, and although they set out the history of how the ET had not originally sent the judgment to her and complained that this was in breach of the ET Rules they did not identify any arguable error of law on the part of Judge Eady. Lewison LJ refused permission to appeal on the papers on that basis.

However, at the oral renewal hearing the Appellant was represented by pro bono counsel (not Mr Margo). Gloster LJ gave permission, albeit with hesitation, because she believed that it was arguable that the original failure by the ET to send the Appellant a copy of its judgment, and then the long delay before it eventually did so, constituted exceptional circumstances. She added that it was also arguable that it was wrong as a matter of principle to attach no weight to the Appellant’s health problems.

On 2 May 2018, i.e. about three weeks before the hearing in this Court, Mr Margo, who had only recently been instructed, notified the Court and the Respondent that he would be applying for permission to amend the grounds of appeal to add a third ground, namely that since the ET had sent the judgment to the wrong address it had not been “sent to the parties”, within the meaning of rule 3 (3), on 28 April 2015 and was only so sent on 4 June, which is 41 days before the date that the appeal to the EAT was filed, with the result that no extension was necessary: I will refer to this as “the in time point”. The Respondent opposed the grant of permission to amend.

There are accordingly three issues on the appeal in Bonnie:

The most well-known statement of this Court’s approach to allowing a point to be taken for the first time on appeal is in Pittalis v Grant [1989] QB 605. (Footnote: ²) Nourse LJ, giving the judgment of the Court, said at p. 611 C-F:

Mr Margo argued that the in time point was indeed one of pure law, as regards which there was no dispute of fact, and that the case would not have been run any differently by the Respondent in the EAT if the point had been taken there, not least because there was recent EAT authority contrary to his case (albeit, he submitted, wrongly decided) by which Judge Eady would have regarded herself as in practice bound – see Carroll, which I discuss at para. 28 below. The Respondent had sufficient notice of the point and was in a position to address it on its merits. That being so, justice plainly favoured allowing it to be taken now. Mr Margo also pointed out that the point was not in fact being considered for the first time. It was apparent from the decision of the Registrar that she understood the Appellant to be arguing that time ran from 4 June: that argument was rejected by reference to Carroll (which may explain why it does not appear to have been raised before Judge Eady). He also said that the fact that the same point arose in Rana showed that it was a point of significance beyond the present case, and since in Rana the Respondents had decided not to appear it would be most unfortunate if the point were decided without the benefit of oral argument on both sides.

Mr Moretto emphasised the late stage at which the point was being sought to be taken. Not only was it not taken before Judge Eady, but it had not been taken in the original Appellant’s Notice or by counsel at the permission hearing before Gloster LJ. More substantially, he contended that if it had been taken before Judge Eady more factual evidence would, or at least might, have been required: the Appellant’s repeated attempts to get a copy of the judgment from the ET were fully supported by the documents produced by her, but there was nothing showing what efforts, if any, she had made to obtain a copy from the Respondent or from Grand & Machyle. He also pointed out that if the point had been taken from the start permission to appeal would (assuming for the present that it was arguable) have been given on the papers, instead of ten months later at the oral renewal: this meant not only delay, which could potentially be prejudicial, depending on how matters went hereafter, but also that the Respondent had wasted costs, which were in practice irrecoverable, in responding in its original skeleton argument to points that were now said not to arise.

I would allow the in time point to be taken. The fact that it is taken late, and could have been taken earlier, is the reason why permission is needed; but it is not in itself a reason for refusing it. I can see no basis for the argument that the appeal would, or even might, have been conducted differently before Judge Eady if the point had been taken at that stage. Even on the basis that the only issue in the EAT was whether time should be extended, it was just as much in the Respondent’s interest to show, if it could, that the Appellant could have obtained the judgment earlier from itself or from her own solicitors, or even that she had in fact done so; and if the point was not explored that was no doubt because the Respondent had no basis on which to challenge her account. In truth, we are here concerned with a pure point of law about the meaning of rule 3 (3). That being so, I do not believe that Mr Moretto’s other points are of any weight. He may be right that if the point had been taken earlier permission would have been granted sooner; but I cannot believe that the additional ten months’ delay, regrettable as it is, is of real significance in the overall time-scale of the appeal or the wider litigation. I do not understand the point about wasted costs: Mr Margo has not abandoned the grounds on which Gloster LJ gave permission.

The issue is whether a judgment is “sent to the parties” where in the case of one of the parties it is sent to a person other than that party or his or her representative. I should make two points by way of preliminary.

First, it was common ground before us, and is plainly right, that it is irrelevant for the purpose of rule 3 (3) when or whether the judgment is received by any party: what matters is when they were sent. That might be thought to be clear from the words of the rule, but it was in any event decided by this Court in Gdynia American Shipping Lines (London) Ltd v Chelminski [2004] EWCA Civ, [2004] ICR 1523, approving earlier decisions of the EAT to the same effect in Hammersmith & Fulham LBC v Ladejobi [1998] UKEAT 500/98, [1999] ICR 673 and Sian v Abbey National plc [2003] UKEAT 0207/03, [2004] ICR 55 (and implicitly also the earlier EAT decision in Mock v Inland Revenue [1999] UKEAT 997/98, [1999] IRLR 785). However, it does not follow that the reasoning in those cases may not have some application to the issue which we have to decide. We were referred in particular to Sian. An important part of Burton P’s reasoning in favour of time running from the date of despatch was that it was inherently more certain than the date of delivery: see para. 15 of his judgment (pp. 61-62). He acknowledged that in a case where the judgment had not come to the attention of a party through no fault of their own or their representative – most obviously, where it had been lost in the post (Footnote: ³) – it might seem unfair that time should be running for a substantial period, and might indeed expire, before they became aware of it; but he held that any such unfairness could be met by the exercise of the discretion under rule 37 – see para. 17 (pp. 62-63). As will appear, analogous arguments were advanced before us.

Secondly, although what happened in this case was that the judgment was sent to a former representative, that is not the only kind of error that could occur. It will be important to test any proposed answer as it would apply to other kinds of “mis-sending”. Two in particular were canvassed in argument, namely:

The only authority to which we were referred which directly addresses the actual issue before us is the decision of the EAT (Judge Hand QC sitting alone) in Carroll v Mayor’s Office for Policing and Crime [2015] UKEAT 203/14, [2015] ICR 835. In that case the claimant/appellant was acting in person, but the tribunal carelessly sent the judgment and written reasons not to him but to a trade union representative who had represented another claimant in linked proceedings: the case was thus in category (1) in para. 26 above. Judge Hand held that the judgment had been “sent to the parties” for the purpose of rule 3 (3). He considered the position in some detail, but I think it is sufficient to summarise his reasoning as follows:

Mr Margo submitted that Carroll was wrongly decided. First and foremost, it reached a conclusion which was contrary to the plain terms of the rule. The passage quoted at (4) above erred by concentrating on “the physical sending out of [the judgment]” and overlooking the words “to the parties”. As a matter of ordinary English, a document cannot be said to have been “sent to the parties” where it has in fact been sent to someone who is not a party or their (current) representative. He submitted that a judgment was only “sent to the parties” when it was sent out in accordance with the relevant ET rules (now rule 86 of the 2013 Rules), i.e. to the parties or any current representative: in his oral submissions he appeared to submit that it was necessary also that it should be sent to the address of which the tribunal had been notified.

As to Judge Hand’s other points, Mr Margo submitted that:

Mr Young’s submissions were to broadly the same effect as Mr Margo’s. However, he disavowed any submission that time would not start to run in a case of “pure” mis-addressing, i.e. my category (2) at para. 26 above. He said that such a case could properly be distinguished from the case where the judgment was sent to the wrong person altogether. It was reasonable to say in the former case that the judgment had indeed been “sent to the parties”, albeit mis-addressed, whereas in the latter case that was simply not possible.

Mr Moretto adopted the reasoning in Carroll. He submitted that it was wrong to read the phrase “to the parties” too literally: taking the language of the rule as a whole, what it connoted was simply that the tribunal should have sent the judgment out, even if there was a mistake in the identity of the addressee or the address. The essence of what the rule was referring to was a formal act of promulgation, analogous to a “hand-down” in the ordinary courts. He did not flinch from the consequence that sending it to someone wholly unconnected with the case would fall within the terms of the rule: any resulting unfairness could be met by the exercise of the rule 37 discretion. He submitted that there was no rational difference between the present case and that discussed in Sian: in both the party in question would be unaware of the promulgation of the judgment through no fault of their own.

However, Mr Moretto made some additional points not made in Carroll. In particular:

The arguments in the Council’s skeleton arguments in Rana were broadly to the same effect as Mr Moretto’s. It did, however, submit that the effect of the EAT Practice Direction was that the date of sending endorsed on the judgment itself should be treated as conclusive.

Mr Margo and Mr Young developed their arguments clearly and persuasively, and they are also powerfully endorsed by McCombe LJ in his judgment. In those circumstances I have not found the issue straightforward. In the end, however, I have come to the conclusion that time did start to run in Bonnie on 28 April notwithstanding the mis-sending of the Appellant’s copy of the judgment to her former solicitors. My reasons are as follows.

The starting-point must be the words of the rule itself. I see the force of Mr Margo’s submission that as a matter of ordinary English it is hard to describe a document as having been “sent to the parties” when in the case of one of them it has been sent neither to the party himself or herself nor to their nominated representative. I also agree with him that the reasoning at para. 52 of Judge Hand’s judgment in Carroll (see para. 26 (4) above) is unconvincing: to say that a document is “sent to the parties” even where it is sent to someone who is not a party only asserts what it is necessary to prove, and Mr Margo’s submission does not require any words to be read in (at least as regards “category (1) cases”). However, I also see force in Mr Moretto’s point that the language of the rule refers to a single date, which is difficult to reconcile with a construction under which different dates would apply in the event of the mis-sending of the judgment to one of the parties. That being so, a strictly literal approach may not be appropriate: the phrase “sent to the parties” could reasonably be read as connoting the act of promulgation – perhaps equivalent to “sent out” or “published” – rather than the mechanics of despatch to each of the parties. I am not to be taking as saying that, as a matter of language, Mr Moretto’s construction is to be preferred to Mr Margo’s – only that both are reasonably possible. But once that point is reached it is necessary to consider which produces the more sensible result in practice.

As to that, I prefer Mr Moretto’s contention. In my view there is an obvious practical advantage in having a single, contemporaneously recorded, date from which time for appealing runs for both parties, so that everyone knows where they stand from the moment that the judgment is promulgated; and I think it very likely that that is what the rule-maker intended. Even though there appears to be no formal obligation on the tribunal to include the standard-form endorsement recording the date at which the judgment is sent to the parties, the EAT Practice Direction clearly shows that that was the practice at the time that the Rules were made, and it can safely be assumed that the rule-maker was aware of it (and even if he was not he would certainly have assumed that the date of sending would appear in the covering letter).

Of course justice requires that where the judgment is mis-sent the party affected should not be unfairly prejudiced by the tribunal’s mistake; but construing the rule so that time will not run until whatever date the mistake is corrected is a rigid solution which goes further than is necessary to correct the problem. As Mr Moretto pointed out, where the tribunal has sent the judgment to the wrong person the victim of the error may in fact receive it (and the covering letter) from the other party, or from the person to whom it has been mis-sent, despite the mis-sending. As from that point they would be substantively in the same position as if the error had not occurred, and there is no good reason why they should not proceed to file the Appellant’s Notice timeously. It would be extraordinary if they could excuse any delay by arguing that time had not started to run until whatever subsequent date the tribunal formally re-sent the judgment to the correct address. It is an important policy of the law in this area that parties should have to decide promptly whether they wish to appeal. Indeed the victim might not even trouble to insist on a formal re-sending at all, in which case presumably they could argue that time had not started to run at all.

In my view a just outcome in mis-sending cases can be achieved more flexibly and sensibly, as Judge Hand held in Carroll, by the appropriate use of the EAT’s discretion to extend time under rule 37. I do not agree with Mr Margo that having to rely on the discretion causes any substantial prejudice to the party in question, at least if the principles governing its exercise are clear. I consider what those principles should be at paras. 42-45 below.

That would be my view even if we were concerned only with cases where the tribunal sends the judgment to the wrong person. But I see force in Mr Moretto’s point summarised at para. 32 (3) above. It is difficult to see a rational basis for distinguishing between “wrong person” and “wrong address” cases, and neither Mr Margo nor Mr Young suggested one.

My view is further reinforced by the consideration that this construction means that cases of mis-sending will be treated in the same way as cases of the kind discussed in Sian, i.e. where the judgment is sent to the correct party and address but goes astray in the post – that is, not by an absolute rule but by the exercise of a judicial discretion (though it does not follow that the way the discretion is exercised will be identical – see para. 49 below). I accept that taking a different approach could be rationally justified, for the reason given by Mr Margo as summarised at para. 29 (2) above. Nevertheless, as Judge Hand observed in Carroll, the basic problem is the same – that is, the party in question has not received the judgment, through no fault of their own – and it is better that the law should adopt the same approach to its resolution.

I would accordingly hold that Carroll is rightly decided, even if I do not adopt every element in Judge Hand’s reasoning, and I would dismiss this ground of appeal.

Cases where the tribunal has failed to send the judgment to the correct party or the correct address are in my view inherently different in character from the familiar run of cases in which a party asks the EAT to exercise its discretion to extend time under rule 37 following some failure on their part. The tribunal has made a mistake, as regards a matter of fundamental importance, and the guiding principle should be that the party affected by that mistake should not be put in a worse position than if it had done its job properly. Since they would in that case have had 42 days from the date of sending to file a notice of appeal, I believe that it should follow – subject to the particular points discussed below – that the discretion should be exercised so as to allow them the same period from the date that they are eventually sent a copy of the judgment, whether from the tribunal or from another source – or (an important qualification) from any earlier date that it would have been sent if they had taken reasonable steps to obtain it. In the absence of some very particular reason to the contrary, it does not seem to me to be fair to allow the victim of a mis-sending less than the full period which the rule-maker judged to be reasonable to allow a party to decide whether to appeal. However, I need to say rather more about some aspects of that approach.

The most straightforward case is where the tribunal is alerted to its mistake and re-sends the judgment, and the standard-form covering letter (which gives information, inter alia, about how to appeal), to the victim of the mis-sending. In such a case the victim is as from that date in the position that they would have been in if they had been correctly sent from the start, and I would expect the discretion to be exercised so as to allow 42 days from the date of re-sending. No doubt the judgment would still in such a case bear the standard-form endorsement recording the date of the original (mis-)sending; but the date of the re-sending should normally be clear from any covering letter or the e-mail to which it was attached.

However, there will also be cases in which the victim receives a copy of the judgment first, or instead, from the other party or from the person to whom it was mis-addressed, rather than from the tribunal. In that situation also they will be in substantially the same position as if it had been correctly sent by the tribunal, and the discretion should be exercised on that basis. That is subject to two possible refinements. First, if the party in question receives the judgment but not the standard-form covering letter, they might be in a less good position than if it had been properly sent, since they would not have received the information about how to appeal. Secondly, since where the judgment is sent by the tribunal by post the party in question will not receive it until at least a couple of days from despatch, on a very strict view it might be thought over-generous to allow the victim 42 days from the date of receipt. I doubt if either point will arise often in practice; but how to deal with them where they do is best left to the discretion of the Registrar or Judge in accordance with the circumstances of the particular case.

That is subject to the qualification referred to in para. 42. There will be cases where the victim of a mis-sending is by one means or another put on notice of what has happened without, however, being sent a copy of the judgment at that point. In such a case it must be incumbent on the victim to take reasonable steps promptly to obtain a copy – most obviously from the tribunal itself, but perhaps also or instead from the other party or the “mis-addressee”. If they fail to do so I do not believe that it would be right for an extension to be granted which put them in a better position than if they had acted with reasonable despatch. There may also be cases where, even in the absence of such notice, so much time has passed since the end of the hearing that a reasonable person would make enquiries of the tribunal to find out what has happened; if the victim only makes such enquiries, or otherwise learns what has happened, after the expiry of such a period it may not be right to extend time in their favour.

I believe that the exercise of discretion in accordance with those principles will be sufficient to address the obvious unfairness that could otherwise result from adopting the construction of rule 3 (3) which I favour. I am conscious that it may seem generous to appellants who have in fact, despite the mis-sending of the judgment, received it only a few days later than they should have, and who nevertheless then take the full period of six weeks (or six weeks less a couple of days) before filing an appeal. But it would be impossible to say how great an abbreviation of the time for appealing is acceptable before it becomes reasonable to grant an extension, and any attempt to calibrate it to the circumstances of the particular case would require a disproportionate examination of the facts and be likely to lead to inconsistent outcomes. It is in my view both more principled and more straightforward to exercise the discretion on the basis that the party affected by the mis-sending is entitled to the full period which the rules intended that they should have.

It goes without saying that the victim of a mis-sending must provide the EAT with the evidence to justify the exercise of discretion in his or her favour. It will be necessary to state not only that the judgment was mis-sent (and, if this is not obvious, how the mis-sending occurred), and was accordingly not received when it should have been, but also to explain the circumstances in which it was eventually received and any steps taken to obtain it in the meantime.

In Carroll Judge Hand did not approach the exercise of the discretion in that way. He directed himself at paras. 35-36 of his judgment (pp. 848-9) in accordance with the standard Abdelghafar guidance and also referred to an observation by Burton P in Sian that the solicitors in that case, once they learnt that the judgment had not arrived, should have acted “with extreme diligence” (see para. 34 (p. 847)). As I have already said, I do not believe that the guidance in the Abdelghafar line of cases is apt to the particular case where the time for appealing provided for in the rules is abbreviated, or missed altogether, as a result of the mis-sending of the judgment by the tribunal. I should say, however, that I do not believe that the result in Carroll would have been any different if Judge Hand had directed himself in accordance with what I believe to be the correct principles. The judgment was “sent to the parties” on 15 August 2013, but, as noted above, the appellant’s copy was sent to the representative of a party in related proceedings. On 9 October his solicitors became aware of what had happened. After various confused communications between them and the tribunal they were finally sent a copy of the judgment on 22 January 2014. The appeal was filed on 3 March (i.e. 38 days later). Judge Hand found that a substantial part of the delay between October 2013 and January 2014 was the fault of the solicitors. In those circumstances I see nothing inappropriate in refusing an extension on the basis that reasonable steps had not been taken promptly to obtain a copy of the judgment; and that was indeed a substantial part of Judge Hand’s reasoning (see paras. 65-66 (pp. 856-7)).

I should make it clear that we are here concerned with cases where the judgment is mis-sent by the tribunal. It should not be assumed that an identical approach will be appropriate where that element is absent and the judgment has gone astray for some other reason which does not involve any fault on the part of the tribunal and may well be undiscoverable. In Sian Burton P refused an extension where the appeal was filed five weeks after the appellant’s belated receipt of the judgment, partly on the basis that he and his advisers, knowing that the primary time limit had expired, should have acted with “extreme diligence”, though he also took into account an earlier delay in chasing the tribunal. I would certainly not wish to suggest that that decision was wrong.

I turn to Judge Eady’s approach in the present case. It is clear from para. 20 of her judgment (see para. 15 above) that she did not believe that she should exercise her discretion so as to allow the Appellant the full period of six weeks from when she was sent the judgment on 4 June. It is implicit in her reasoning that she believed that the Appellant should, since the primary time limit had expired, have acted with particular expedition. For the reasons which I have given, I believe that this was the wrong approach. I believe that she should have been prepared to grant an extension to the date that the appeal was filed, which was 41 days from 4 June, subject only to the question whether the Appellant should reasonably have obtained a copy of the judgment earlier. There is, however, no room for criticism of the Appellant on that score: she wrote promptly and persistently to the tribunal to ask for the judgment, and the delay of five weeks in her being sent it was entirely the tribunal’s fault. In my view the only proper exercise of the Judge’s discretion was to grant the extension sought, and I would allow the appeal accordingly.

It is, again, not necessary to say anything about the substance of the Appellant’s claims, which were, broadly, of sex discrimination and whistleblower detriment. Although she named as respondents not only her employer, the London Borough of Ealing, but also one of its employees, for present purposes I need not distinguish between them and will refer simply to “the Council”.

The Appellant was initially represented by a firm of solicitors called Seth Lovis & Co; but in May 2014 they informed the Council and the tribunal that they were no longer acting.

The hearing of the Appellant’s claims began in the London Central Employment Tribunal on 1 September 2014. She was represented by counsel instructed via direct access. At the conclusion of the hearing, on 8 September, the tribunal announced its decision dismissing the claim and gave oral reasons. The following day Employment Judge Tayler signed a formal judgment, which is endorsed as having been sent to the parties on 12 September.

The Appellant’s copy of the judgment was sent not to her in person but to Seth Lovis, who were no longer acting. The lawyer responsible for the case in the Council’s legal department, Mr Quill, who behaved with admirable punctiliousness throughout, noticed that it had been mis-sent and notified the tribunal, the Appellant and her counsel by e-mail. He attached to his e-mail a scanned copy of the judgment and tribunal’s standard-form covering letter.

Under rule 62 (3) of the ET Rules of Procedure the Appellant was entitled to request written reasons “within 14 days of the sending of the written record of the decision”. She made such a request by e-mail dated 24 September 2014: that was within the 14-day period, so the fact that the judgment had been mis-sent is immaterial. In her e-mail she asked that the reasons be sent to her home address. She noted that she understood that she would have fourteen days to appeal once she had received the reasons: that was incorrect, since the period is 42 days, but the error does not seem to have had any material consequences.

The written reasons were signed by Employment Judge Tayler on 4 November 2014 and endorsed as having been sent to the parties on 6 November. They were received by the Council on 10 November. Mr Quill noted that it was unclear from the face of the covering letter whether they had been sent to the Appellant personally or, again, to Seth Lovis; and he accordingly e-mailed her the same day to apprise her of their receipt and suggest that if she had not received a copy she should contact the tribunal. He did not, however, on this occasion attach a copy. She had indeed not received the reasons. She forwarded Mr Quill’s e-mail to the tribunal on 11 November and again (giving the case reference) on 12 November, asking for the reasons to be sent to her home address. So far as appears from the papers before us, the tribunal failed to reply. On 14 November she e-mailed again, saying that if they had not been received in the meantime she would attend at the tribunal in person to get a copy. She did so on 17 November and was given a copy of the reasons, though not of the covering letter that would have accompanied them as sent.

If time ran from 6 November 2014, the Appellant’s time for appealing expired on 18 December. The Appellant filed a notice of appeal at the EAT on 17 December, together with the ET1, the ET3, the judgment and the written reasons, all of which are required by rule 3 (1) of the 1993 Rules to be filed with the notice of appeal. Unfortunately, however, she did not include the grounds of claim and grounds of resistance which formed part of the ET1 and ET3 respectively, and there was also a page missing from the reasons. She was notified of these defects by e-mail the same day; but she did not look at her e-mails till 19 December. She supplied the missing documents on 22 December, which was the next working day, but that was four days out of time.

The Appellant applied for an extension. The application was dismissed by the Registrar. She appealed from that decision. At the hearing before Judge Eady, which was on 9 June 2015, she represented herself.

Judge Eady begins the section of her judgment where she gives her reasons for dismissing the appeal by saying (para. 18) that “I have no reason not to accept what the Claimant has told me”. She does not spell out precisely what the account which she was accepting was; but she had previously, at para. 16, recorded the Appellant as saying that “the ET sent the written reasons to the wrong address”, and that she had not received them until she went to collect them from the tribunal. I accordingly read her as accepting that the reasons had indeed been sent to Seth Lovis and not to the Appellant herself. There was no documentary evidence before her unequivocally demonstrating that that was the case, but it was a reasonable inference, given what had happened with the judgment: one would like to think that the tribunal would not make the same mistake twice, particularly as the Appellant had expressly asked for the reasons to be sent to her home address, but the fact is that they were not received by her, and it is not difficult to imagine how a tribunal official, sending out the reasons several weeks later, simply consulted the (still uncorrected) details on the database and overlooked the subsequent correspondence. Ultimately, however, all that matters is that it is in my view adequately clear that Judge Eady accepted that the reasons had been sent to Seth Lovis and not to the Appellant.

Having thus accepted the Appellant’s factual account, Judge Eady observed, however, that “that does not properly explain the default”. She continued (still in para. 18):

On that basis, Judge Eady approached the case, conventionally, as one where the Appellant had left it to the last moment to file her notice of appeal and had made an error which it was then too late to correct. At paras. 19-21 she considers the Appellant’s explanation for leaving the filing of the appeal so late and for omitting the grounds of claim and the grounds of resistance and finds them inadequate. For reasons that will appear I need not give the details of her reasoning on this aspect.

The Appellant’s grounds of appeal are in some respects a little obscure, but they squarely take the point that the EAT should have treated time as running not from 6 November, when the reasons are recorded as having been sent to the parties, but from 17 November, when she collected her copy from the tribunal: on that basis 22 December, when she filed the missing documents, was within the 42-day period under rule 3 (3), and no extension was necessary. The Appellant asked permission to rely on fresh evidence which she said clearly established that the reasons had indeed been sent not to her but to Seth Lovis.

Laws LJ refused permission on the papers, stating succinctly that “time runs from the date when the written reasons are sent, even if addressed to the wrong person”. But at the renewed oral permission hearing on 27 October 2016, where the Appellant was represented by pro bono counsel (not Mr Young), Longmore LJ gave permission on two grounds. The first was that time did indeed run only from 17 November: he acknowledged that Carroll was authority to the contrary, but he said that it was arguable that it was wrongly decided. The second ground was that, even if time ran from 6 November, Judge Eady had erred by not exercising her discretion to extend time. He adjourned the hearing of the Appellant’s application to adduce new evidence to the hearing of the substantive appeal.

It is not in fact clear that the Appellant advanced before Judge Eady the first of the grounds permitted by Longmore LJ: certainly she does not expressly address it in her judgment, though that may be because she regarded herself as in practice bound by Carroll. But in its full and well-argued skeleton argument dated 8 February 2017 the Council does not object to the point being taken and addresses it on the merits. It does, however, in a supplementary skeleton argument dated 30 November 2017, object to the admission of the fresh evidence.

Accordingly the substantive issues before us are:

The evidence on which the Appellant wishes to rely consists of three documents which are said to constitute or record confirmations by the ET itself that on 6 November 2014 the reasons had been sent to Seth Lovis and not to her personally. The nature of the individual documents is a little complicated, but for reasons that will appear I need not go into the details.

The Council opposes the application on the basis (a) that if Carroll is correctly decided the evidence is irrelevant, and (b) that, if for any reason it nevertheless remains relevant, the issue is one of fact that ought to be remitted to the EAT. I have already held that Carroll was correctly decided on this point, and it follows, as the Council contends, that the appeal was out of time whether or not the reasons were sent to the wrong person. I do not, however, accept that it follows that the question whether they were in fact mis-sent is irrelevant: as I have explained in Bonnie, where the tribunal itself has sent the judgment or reasons to the wrong party or address, a particular approach is required to the exercise of the discretion to extend, and it is accordingly necessary to establish whether that did indeed occur. The fresh evidence is therefore in principle potentially relevant. However, I do not believe that in this case its admission is necessary for that purpose. For the reasons given at para. 59 above, I believe that Judge Eady accepted that the reasons, like the judgment, were sent to Seth Lovis and not to the Appellant personally, and there is no need for further evidence to support a finding already made.

I would accordingly refuse the application, on the basis that the Appellant can establish the relevant part of her case without it. For what it is worth, however, we did look at the documents de bene esse; and, although the story is complicated by what appears to have been a well-intentioned but misguided attempt by the ET to correct the record retrospectively, they do in my view confirm that the reasons were sent to Seth Lovis. (Footnote: ⁶)

It is clear that, as in Bonnie, Judge Eady did not approach the exercise of her discretion on the basis that the Appellant ought to be given the same period that she would have been entitled to if the tribunal had sent the reasons to the correct address: she says so in terms in the second sentence of the passage from para. 18 of her judgment quoted at para. 60 above. In this case too there is no question of any culpable delay on the Appellant’s part in obtaining a copy of the reasons once she was alerted to the fact that they had been mis-sent. Accordingly it follows from the principles which I have discussed in relation to Bonnie that the only proper exercise of the discretion was to grant an extension until 22 December, being the date at which the appeal was properly constituted, which was of course within 42 days from when she first received the reasons.

It is clear that what weighed particularly with Judge Eady was the fact that the Appellant herself at the time appears to have understood that time ran from 6 November and tried to file her appeal in time on that basis, failing only because of the missing documents. Her point at the start of para. 18 of her judgment (para. 59 above) is, in effect, that the mis-sending of the reasons was not the real cause of, or explanation for, her default. I take the point, and in another context I accept that such a point could be a good reason for refusing to exercise the discretion in an appellant’s favour. But I do not think that it has that effect here. If the underlying principle is that the discretion should be exercised so that the party in question has the same time to appeal from when the judgment or reasons are in fact sent as if they had been correctly sent in the first place, then it makes no difference that he or she is in fact aiming to meet the original deadline.

I would accordingly in this case also allow the appeal and grant the extension necessary to bring the Appellant’s appeal within time.

The disagreement within the Court about the correct route to be followed does not affect the result, and our judgments already promulgated provide, as I have said, for both appeals be allowed and both cases to be remitted to the EAT for consideration under the “sift” process. I should emphasise that our reasoning has involved no consideration of the substantive merits of the proposed appeals, and it remains to be seen whether the EAT will find that they raise any arguable point of law. It is regrettable that it will have taken so long for the appeals to reach that preliminary stage; and still more so if they are permitted to proceed and are eventually successful. That is, I fear, largely a result of the extreme listing pressures in this Court (now fortunately somewhat abating); but I am sorry that the delay has been increased by our not being able to reach a decision in these appeals as soon after the hearing as we would have wished.

I agree with Underhill LJ that both these appeals should be allowed. I agree that, if necessary, an extension of time should be granted to each appellant, if, as my Lord would hold, the appeals by each appellant from the Employment Tribunal (“ET”) were not lodged with the Employment Appeal Tribunal (“EAT”) within the time prescribed by rule 3(3) of the EAT Rules 1993 (as amended).

In my judgment, however, both appeals were in fact lodged in time and, therefore, an extension of time is not required in either case. In other words, I respectfully disagree with Underhill LJ as to the application of rule 3(3) to the facts of these cases.

As Underhill LJ has most helpfully set out, in comprehensive form, the facts and competing arguments in the case, I can keep this judgment, setting out my disagreement with him on the point in issue, quite short.

In essence, I find myself in entire agreement with the argument of Mr Margo in Ms Bonnie’s appeal (supported by Mr Young for Ms Rana), as summarised in paragraphs 28 and 29 of Underhill LJ’s judgment. In my view, Carroll was indeed wrongly decided.

The short point is that to say that the judgment was “sent” to Ms Bonnie or to Ms Rana is, in my view, a straightforward misuse of language on the facts of these cases. One does not “send” something to “John Doe” by sending it to “Richard Roe”. One does not “send” a document to a party to litigation by sending it to the representative of another party (Carroll). Nor does one send such a document to a party by sending such a document to a former representative of a party who (to one’s knowledge) is no longer acting for that party. No ordinary, sensible member of the public would answer in the affirmative the question, “Can I send you an important document by sending it to Father Christmas instead?”

It seems to me to be wrong to say one sends something to someone by sending it to someone else. The only authority that I can think of for such a proposition is that identified by Lord Atkin in Liversidge v Anderson [1942] AC 206 at 245, which I will not repeat; it is well-known.

The reference in that case to words meaning what one chooses them to mean, as Lord Atkin described the effect of the rival meaning in Liversidge’s case is, I think, of real significance here. This is because the respondents’ arguments focus upon possibly awkward consequences, in some individual cases but not in the vast majority of them, of giving the words of this rule their obviously natural meaning.

The desire for clarity in a date of sending is invoked, as is the possibility of there being different dates for “sending” to different parties; there is the possibility of misaddressing a document and the difficulty occurring if the relevant party, to whom the document has not actually been “sent” at all has received a copy of the document from some other source.

As Underhill LJ notes in paragraph 27(4) above, Judge Eady in this case considered that the appellants’ argument would require the rule to be read as if it required the document to be sent “properly addressed”, whereas the respondents’ arguments did not require the reading in of extra words. In my judgment, no such additional words are required if one uses the words of the rule without, as the judge did, failing to give any effect to the express requirement of the rule that the document be sent “to the parties”. The judge’s construction excises words from the rule.

Judge Eady said that,

In my view, the requirement should be taken to be that the document should be sent to the correct party at the address communicated to the tribunal as being his or her address for service. It is the easiest of tasks to require a party to inform the tribunal of that address at the outset and to put the burden upon him or her to communicate any change – as indeed these appellants did. Thereafter, it is sensibly incumbent upon the tribunal to “get it right” and to send it to the up-to-date address supplied which the tribunals here (and particularly in Ms Bonnie’s case) lamentably failed to do. If a document actually arrives at the correct destination, slightly misaddressed, I doubt that it would be found not to have been “sent” to the party on the date so despatched. A sensible practice would be to advise parties to enquire of the tribunal if they have not received the necessary document within a specified period of x weeks. It is to be hoped that that would flush out any error in despatch of the document which may have occurred, even if the experience of Ms Bonnie’s case might indicate otherwise.

Further, in my judgment, it is a misstatement of the problem in the present two cases to say that a judgment is “still sent to a party even if it is sent to…somebody incorrectly believed to be an agent for that party” (as per the judge). In this case, the ET had no reason to believe that the agent to whom the document was sent was the party’s agent; it had been expressly informed that the agent in question was no longer her agent. If it held its belief incorrectly, it had no excuse for so doing.

Underhill LJ says in paragraph 36 above that,

In my judgment, the words of this rule are quite clear. They require the judgment and written reasons to be “sent to the parties”. That is not achieved by sending them to only one of them. Problems only arise when the tribunals do not do what the rules require of them, namely to send the papers to the parties. It is a simple task to do that and, in my judgment, courts should not avoid the consequences of a tribunal failing to perform that task by the artificial device of misreading the words of the rule or misapplying the words to the facts of an individual case.

The helpful guidance which Underhill LJ gives as to the circumstances in which an extension of time ought to be granted would serve to avoid the injustice done below to these appellants in the present cases. However, I do not consider that such guidance, going only to the exercise of the court’s discretion, is any proper substitute for a direct application of the words of the rule themselves.

For these reasons, I would allow both these appeals, but if my understanding of the rule does not find favour with either of my Lords, I would grant the extension of time sought by each of these appellants for the reasons given by Underhill LJ.

Underhill and McCombe LJJ are agreed that both the appeals before us should be allowed and the cases remitted to the Employment Appeal Tribunal for the respective Claimants’ appeals to be considered on its merits. But they differ in the routes by which they reach that destination.

The critical question is one of construction of Rule 3(3)(a)(i) of the EAT Rules 1993 (as amended). This gives the right to institute an appeal from an employment tribunal to the EAT within a period of 42 days “from the date on which the written reasons were sent to the parties”. McCombe LJ holds that the document containing the written reasons is only “sent to the parties” if it is sent to each party, or at any rate the relevant party, at the address communicated to the tribunal by or on behalf of that party as being his or her address for service. Underhill LJ, however, holds that a document is “sent to the parties” if the tribunal sends copies out addressed to each party, even if the address is wrong, or even if the document is sent to an addressee who does not have authority to receive it.

The literal construction accepted by McCombe LJ has the attraction of clarity and simplicity, and of conforming to the natural meaning of the words. But I do not think that it was the result which the draftsman of Rule 3(3) was seeking to achieve. What is plain is that an appeal to the EAT may be lodged as of right within 42 days of the document being “sent to the parties”. If by an error of the employment tribunal staff no such copy is sent to a party who might wish to appeal, time would presumably never start to run at all: a most undesirable outcome. If by a lesser error of the tribunal staff a correctly addressed copy is sent to party A on one day and to party B a week later, then the literal construction breaks down altogether, since there is no single date on which the document is “sent to the parties”.

For the reasons given by Underhill LJ at paragraphs 35-36 of his judgment I agree that it is possible to construe “sent to the parties” as meaning “promulgated” (even if the tribunal then fails to send the document to one party at the correct address). I agree that this construction of Rule 3(3), when coupled with the discretion to extend time conferred by Rule 37, produces the more sensible result in practice; and therefore it seems to me that it was the result which the draftsman was seeking to achieve. I also agree with all that Underhill LJ says about the way in which the Rule 37 discretion should be exercised in cases of this kind where the tribunal has made an error.

It follows that I too would hold that in each of these cases an extension of time in which to appeal to the EAT was necessary; that such an extension should now be granted; and that the appeals should be allowed accordingly.