England and Wales Court of Appeal (Civil Division) Decisions

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM Upper Tribunal (Immigration and Asylum Chamber)
UTJJ Allen & Jackson

		Royal Courts of Justice Strand, London, WC2A 2LL
		21/02/2019

Lord Justice Underhill:

INTRODUCTION

1. This is a preliminary hearing in a pending appeal against the decision of the Upper Tribunal in a Country Guidance case about the return of Afghan nationals to Kabul – AS (Safety of Kabul) Afghanistan CG [2018] UKUT 0018 (IAC) (UTJJ Allen and Jackson). It raises a question about the circumstances in which the Tribunal can correct errors in the reasons which it gives for its decisions. The background is as follows.
2. One of the issues in the Tribunal was the likelihood of residents of Kabul suffering death or injury in "security incidents". There was evidence before it about both the number of such incidents and the population of Kabul. In a speaking note submitted to the Tribunal as the basis of his closing submissions (which were on 20 November 2017) Mr Sarabjit Singh QC, for the Secretary of State, said (at para. 64):

"Based on the UNAMA 2017 Mid-Year Report, there may be 2,100 civilian casualties in Kabul province this year (as the mid-year figure is 1,048 …). Even if the 'real' figure is more like 5,000 casualties, that is still only about 0.1% of the population of Kabul province (if the population is 4.5m). That means that 99.9% of the population of Kabul will not be casualties of indiscriminate violence in Kabul this year. Security situation should be viewed in that context."

(In fact 5,000 is, more precisely, 0.111% of 4,500,000, but nothing turns on that for our purposes.)

3. The Tribunal did not promulgate its decision until 23 March 2018. I should note at this stage that a draft was not pre-circulated to counsel, as it would have been had it been a reserved judgment of the High Court, and there was accordingly no opportunity for them to draw any errors to its attention. I return to this point below.
4. In the course of the recitation of the submissions in its decision the Tribunal said, at para. 106:

"… Population estimates for Kabul city range between 3.5 and 7 million people. The Respondent, assuming that the total civilian casualties for 2017 based on figures from the UNAMA mid-year report are 2,100, submits that that shows a very low percentage of the population affected. Even if there were 5000 civilian casualties in a year, based on a population of 4.5 million, that still equates to less than 0.01% of the population affected."

When it came to deal with the issue in its actual reasoning, it said (at para. 196):

"However, despite the number and impact of security incidents in and around Kabul city, we find that these are not at such a high level so as to make internal relocation to Kabul unsafe. In particular, although not necessary to reach the threshold in Article 15(c) of the Qualification Directive, we note that the evidence before us shows that the level of indiscriminate violence falls very far short of that sort of threshold and directly affects (by way of death or injury) only a tiny proportion of the population of Kabul city - less than 0.01% even if there were 5000 incidents in a year (more than double the numbers recorded by UNAMA in the first half of 2017 plus the same again assuming the same numbers in the second half of 2017) with a population of 4.5 million. The calculations vary depending on the population estimates but even on conservative calculations with high casualty figures and low population estimates, the percentages of people affected are incredibly small. This remains the case even with significant underreporting of casualty figures by UNAMA based on their strict methodology for casualties to be included."

5. It will be seen that although the first of those passages is clearly intended to reflect the submission made in Mr Singh's speaking note, and the second (in substance) to accept that submission, they do not reproduce his percentages. The most striking difference is that the Tribunal uses a figure of 0.01% (that is, 1 in 10,000), whereas Mr Singh's figure was 0.1% (1 in 1,000); but it should be noted also that it says "less than" instead of "about". Not only does the Tribunal's quantification of the risk not reflect the submission made, but it is common ground before us that there is no support for it in the evidence and that it is simply an error. To anticipate, the question which gives rise to the present hearing is whether the Tribunal had regard to the erroneous quantification in its actual reasoning or whether it is simply an accidental slip (whether or not literally typographical) in the expression of reasoning based on the correct quantification: I will refer to the former alternative as "a substantive error" and to the latter as "an error of expression".
6. Following the promulgation of the Tribunal's decision the Appellant applied to it for permission to appeal on nine grounds. Ground 1 was that:

"In concluding that internal relocation to Kabul is both 'available' and reasonable under the Refugee Convention despite the level of insecurity and violence in the city [paras.190-199], the UT made material errors of law."

Five particular errors were then pleaded, one of which – (c) – was:

"In any event the UT made a serious error of fact, amounting to a material error of law [para. 196]: - they erred as to the calculation of death/injury relative to population, on which finding it placed significant weight in reaching its conclusion on reasonableness."

7. That pleading was evidently intended to rely on the error explained at para. 5 above, but, most regrettably, it does not say what the error was; and it is clear that the Tribunal, entirely understandably, failed to appreciate what was being asserted. In its reasons for refusing to give permission on this point (which was on the papers and not at a hearing) it said:

"The Applicant does not identify how the UT erred in fact as to the calculation of the risk of death/injury relative to population explained in paragraph 196, by reference to the evidence before it or otherwise."

8. The Appellant filed an Appellant's Notice with this Court. The pleaded grounds were now limited to six and differently expressed. Ground 1 reads:

"The UT found that relocation to Kabul would be reasonable notwithstanding the level of insecurity and violence in the city. In considering that issue, and reaching that conclusion, the UT made a material mistake of fact, amounting to an error of law, namely:

The UT made a serious error fact, amounting to a material error of law [para. 196] – it made a simple but very significant miscalculation regarding risk of death or injury relevant to population, a calculation to which the UT attached primary importance."

9. Despite the UT's comment on the original version, that pleading still fails to say what the "significant miscalculation" actually was. This Court has repeatedly said that grounds of appeal should be succinct, but that does not mean that where a specific point is being made it should not be at least identified. In this case, however, unlike the application to the Tribunal, the Applicant had to file a skeleton argument; and that did squarely identify the 0.01%-for-0.1% error.
10. In accordance with the usual practice in Country Guidance cases the application for permission to appeal was listed for an oral hearing inter partes. That took place before me on 28 November 2018.
11. It was Mr Singh's case at the hearing that permission should be refused on ground 1 because the reference to 0.01% was clearly a typographical error. The Tribunal had had the correct percentage, with the underlying figures, in his speaking note and it was inconceivable that it had regard to any different percentage in its deliberations: there could be no reason why it should have ignored the figure that it was given and have carried out its own calculations. It was clear that its substantive reasoning was based on the correct figure and there had simply been an accidental error of expression.
12. Ms Sonali Naik QC, for the Appellant, submitted that the use of the 0.01% figure could not be assumed to be merely an error of expression. It appeared in two separate passages, and the Tribunal departed from the speaking note not only by inserting a zero but also by substituting "less than" for "about". I return to this part of her submissions in more detail at para. 42 below.
13. I raised with the parties whether it was legitimate to resolve the problem by asking the Tribunal itself to clarify, if it could, the nature of the error. I pointed out that if the draft had been pre-circulated, or if the error had been properly identified in the grounds of appeal, the Tribunal would certainly have responded to it, either by saying that it was indeed a mere error of expression (typographical or otherwise) or by acknowledging that it had worked on the wrong figure^[1]; and it would be very unsatisfactory if the failure to take either of those opportunities meant that it was too late to obtain such a response. Mr Singh agreed, but Ms Naik said that it would be wrong in principle for any approach now to be made to the Tribunal. Neither party was in a position to argue the point fully, and I accordingly directed the present hearing in order to resolve the question.
14. There are essentially two questions for us:

(A) Is there a procedure by which the Upper Tribunal can be asked to resolve the question whether the error was substantive or was only an error of expression ?

(B) If there is, should it be employed in the present case ?

I take the two questions in turn.

(A) IS THERE AN AVAILABLE PROCEDURE ?

15. At the previous hearing, when we were considering the available procedures, I floated the possibility that this Court had the power to ask the Upper Tribunal the relevant question under the procedure endorsed by this Court, as regards appeals from the High Court, in English v Emery Reimbold [2002] EWCA Civ 605, [2002] 1 WLR 2409, and again in the context of appeals from an employment tribunal in Barke v Seetec Business Technology Centre Ltd [2005] EWCA Civ 578, [2005] ICR 1373 – "the English/Barke jurisdiction". In her skeleton argument for this hearing Ms Naik submitted that that jurisdiction was unavailable in the context of appeals from the Upper Tribunal, relying on the decision of this Court in Hatungimana v Secretary of State for the Home Department [2006] EWCA Civ 231 (which was, however, based on the Asylum and Immigration Tribunal (Procedure) Rules 2005 and not the Upper Tribunal Rules) and also on the provisions of CPR 52.20, which prescribes the powers of an appeal court.
16. However, it is unnecessary to resolve that question because Mr Singh disavowed reliance on any power enjoyed by this Court to direct, or ask, the Upper Tribunal to amplify or clarify its reasoning. He made it clear that he relied only on the power of the Tribunal itself to correct errors in its decisions under the slip rule embodied in rule 42 of the Tribunal Procedure (Upper Tribunal) Rules 2008 ("the Rules") – or, if that were inapplicable, an analogous inherent jurisdiction. That power could be exercised by the Tribunal itself, either of its own motion or in response to an invitation from a party, without any involvement at all by this Court of the kind which occurs in English/Barke cases. Ms Naik disputed that the Tribunal had any such power, either under rule 42 or under any inherent jurisdiction, in the circumstances of the present case.
17. Accordingly, the essential question under this head is whether the Upper Tribunal has the jurisdiction now relied on by Mr Singh. It was common ground before us that that is a question on which this Court can and should rule even though, if Mr Singh is right, no order from us is required and the Secretary of State could approach the Tribunal direct: it is important that the Tribunal should know definitively whether it has jurisdiction to respond to any such invitation. I will set out the parties' submissions as regards (a) rule 42 and (b) any inherent jurisdiction.

Rule 42

18. Rule 42 falls under Part 7 of the Rules, which is headed "Correcting, setting aside, reviewing and appealing decisions of the Upper Tribunal". It reads:

"The Upper Tribunal may at any time correct any clerical mistake or other accidental slip or omission in a decision or record of a decision by—

(a)   sending notification of the amended decision, or a copy of the amended record, to all parties; and

(b)   making any necessary amendment to any information published in relation to the decision or record."

It was common ground before us that that rule, in common with slip rules applying to other jurisdictions, applies only to what I have above called "errors of expression". If the error was substantial the rule has no application. (That is of course only a shorthand. There is a good deal of law about where the distinction lies which it is not necessary to review here. The two authorities to which we were referred are R v Cripps, ex p Muldoon [1984] 1 QB 686 and Bristol-Myers Company v. Baker Norton Pharmaceuticals Inc EWCA Civ 414, [2001] RPC 45, but there are others.)

19. The Rules were made by the Tribunal Procedure Committee ("the TPC") under powers conferred by various provisions of the 2007 Act. The relevant provision for present purposes is paragraph 15 of Schedule 5 to the Act, which reads, so far as material:

"(1)  Rules may make provision for the correction of accidental errors in a decision or record of a decision.

(2)  …

(3)  Sub-paragraphs (1) and (2) shall not be taken to prejudice, or to be prejudiced by, any power to correct errors or set aside decisions that is exercisable apart from rules made by virtue of those sub-paragraphs."

20. It is convenient to start with Ms Naik's submissions on this aspect. She drew our attention to Part 6 of the Rules, headed "Decisions". Rule 40, also entitled "Decisions", sets out the ways in which the Upper Tribunal may give a decision. Paragraph (1) provides that a decision may be given orally. Paragraph (2) (a) provides that, subject to immaterial exceptions, as soon as reasonably practicable after making a final decision is made the Tribunal must provide the parties with a "decision notice". Paragraph (3) requires it, subject again to immaterial exceptions, "to provide written reasons for its decision with a decision notice provided under paragraph (2) (a)". It is common ground that the decision notice under paragraph (2) is intended to contain only the statement of the decision itself and is distinct from the "written reasons" under paragraph (3). That is illustrated by the document promulgated by the Tribunal in the present case. It is headed "Decision and Reasons". The bulk of the document contains the Tribunal's written reasons, over some 250 paragraphs, but it concludes, on the final page, with a formal "Notice of Decision" which, after reciting an earlier decision to set aside the decision of the First-tier Tribunal, proceeds:

"The decision is re-made. The Appellant's appeal is dismissed on asylum, humanitarian protection and human rights grounds."

21. On the basis of that distinction, Ms Naik submitted that the phrase "decision or record of a decision" in rule 42 is intended to apply only to the decision notice and not to the written reasons. In that case it would not apply to the error in the present case, which is in the Tribunal's reasons and not in the decision notice.
22. In support of that narrow approach to the construction of the phrase "decision or record of decision" Ms Naik relied on the decision of this Court in In re A (a Child) [2014] EWCA Civ 871, [2014] 1 WLR 4453. That case concerned the refusal of a judge sitting in the Family Division of the High Court to correct what were said to be errors in his factual findings in a judgment. The appellant relied on the slip rule embodied in CPR 40.12 (1), which reads:

"The court may at any time correct an accidental slip or omission in a judgment or order."

He submitted that the errors in question formed part of the Judge's "judgment" within the meaning of the rule. That submission was rejected. Patten LJ, with whom Black LJ and I agreed, said, at para. 20 of his judgment (p. 4457F), that "a distinction … has to be made between the judgment or order of the court below and the reasons given for that judgment or order". He relied on well-established case-law about the meaning of the phrase "judgment or order", in contexts other than the slip rule, in the Rules of the Supreme Court ("the RSC", being the predecessor to the Civil Procedure Rules) and in the primary legislation; and specifically on the earlier decision of this Court in Lake v Lake [1955] P 336 (which itself relies on earlier authority). On the basis of that authority he held that the words "judgment or order" in CPR 40. 12 (1) were limited to the "formal judgment or order which is drawn up and disposes of the proceedings". Black LJ pointed out at paras. 38-40 of her concurring judgment (p. 4463 B-E) that the term "judgment" is ambiguous but that in the context of the slip rule it meant only the formal "end product of the proceedings". Ms Naik submitted that the draftsman of the 2007 Act, albeit using different terminology because the tribunals established by it do not give "judgments", must be taken to have been intending to achieve the same result.

23. Mr Singh submitted that rule 42 was not intended to track the distinction in Part 6 between "decision notice" and "written reasons". The phrase "decision or record of decision" was taken directly from the terms of paragraph 15 of Schedule 5 to the 2007 Act, which applied to the rules to be made for the First-tier Tribunal as well as the Upper Tribunal, and there was thus no reason to suppose that it was intended to reflect the different language used in a different part of the Rules. The term "decision" was entirely general, and it was natural to describe the entirety of the document promulgated by the Tribunal as a "decision". The case-law about the interpretation of CPR 40.12 (1), which was differently worded, was of no assistance.

(2) Inherent Jurisdiction

24. Mr Singh referred us to Hazeltine Corporation v International Computers Ltd [1980] FSR 521. In that case Whitford J had delivered judgment in a patent case. In the transcript of the judgment as approved by him a "not" was accidentally omitted in a passage which was potentially material to a pending appeal, and the appellant applied to him to make a formal correction to the transcript. He held that he had no power to make the correction under the slip rule (then embodied in RSC O. 20 r. 8, which, like CPR 40. 12 (1), referred only to correcting "judgments or orders"), because the error was in his reasons rather than his decision: Lake v Lake was not cited to him but other authorities to similar effect were. However, he held that he had an inherent jurisdiction of essentially the same extent as regards the reasons for a judgment. He said, at p. 524:

"To my mind in an appropriate case there must be an inherent power to correct a clerical error in a judgment extending to the reasoning behind the decision or to correct what can properly be described as an accidental slip, but no more than that. While therefore I am not of the opinion that I have any power to do what is sought under [the Rules], the matter having been brought before me I think it only right that I should make the correction sought under what I consider to be my inherent jurisdiction, and I shall do so."

25. Mr Singh submitted that if the High Court enjoyed such a jurisdiction there was no reason why the Upper Tribunal should not do so. It is itself a superior court of record (see section 3 (5) of the 2007 Act); and paragraph 15 (3) of Schedule 5 of the Act expressly preserves "any power to correct errors … that is exercisable apart from rules made under [the powers conferred]".
26. Ms Naik submitted that the power under rule 42 to correct errors in a decision or record of a decision was clearly intended to be exhaustive and that there was no room for a parallel power under the inherent jurisdiction: she referred to R (Kumar) v Secretary of State for Constitutional Affairs [2006] EWCA Civ 990, [2007] 1 WLR 536.

Discussion and Conclusion

27. I should say by way of preliminary that terminology in this area is rather slippery. I have already noted the ambiguity in the term "judgment", but there is a similar ambiguity in "decision", which may refer, depending on the context, either to what Black LJ in A (a Child) calls "the end product of the proceedings" or to the reasons which the Court gives for arriving at that end product. I will call the former "the formal decision" and the latter "the reasons"; and where it is unnecessary to distinguish I will refer simply to "the decision". The distinction is conceptually clear from the authorities, though it may not be straightforward to apply in particular cases (see, for example, Compagnie Noga d'Importation et d'Exportation SA v Australia & New Zealand Banking Group Ltd [2002] EWCA Civ 1142, [2003] 1 WLR 307, discussed in A (a Child); also B (a Minor) [2000] 1 WLR 790).
28. I start by saying that in my view Mr Singh is right that rule 42 cannot be taken as being specifically intended to track the distinctions in Part 6 of the Rules between "decision notice" and "written reasons". The language is different and is derived directly from the statute. The draftsman of the statute must be taken to have recognised that not only the First-tier Tribunal and the Upper Tribunal but the various Chambers, exercising jurisdictions of very different characters and with very different origins, might promulgate their decisions in different ways; there might also be differences in practice and terminology between England and Wales on the one hand and Scotland (and sometimes Northern Ireland) on the other. It must accordingly have been intended that the language of paragraph 15 of Schedule 5 should be of a general character and wide enough to cover decisions given in different formats.^[2] It must have been contemplated that in the case of some kinds of decision both the formal decision and the reasons would be contained in a single undifferentiated document.^[3] If that was the intention of the statute, the same meaning should be adopted when that language is transposed to the Rules made under it.
29. I accept, nevertheless, that there is a conceptual difference between the formal decision and the reasons for that decision, even if it is not always reflected in the form of the decision. The real question is whether Parliament in enacting the provision empowering the TPC, or the TPC in making the rule, intended that the Upper Tribunal (and, in the case of the statute, the First-tier Tribunal) should only have the power to correct slips in its formal decision and not the reasons. I start with the language itself, ignoring at this stage any consideration of the scope of the slip rule in the CPR. As to that, the language is not explicit, and I think there are two possible approaches to the formula "a decision or record of a decision"; but on neither approach do I think that there can be detected a clear intention to limit the power of correction to the formal decision. I take the two approaches in turn.
30. A strict approach would be that it was necessary to give some different content to the two different elements – "decision" and "record of a decision". That can be done by treating the former as referring to the reasons and the latter as referring to the formal decision. I accept that if that were the intention the language is not very apt; but it is not impossible, and I am unable to see any other way of differentiating between them.
31. The alternative approach, which I would favour, is that the strict approach attributes to the draftsman a greater degree of precision than was intended, and that the two elements in the phrase can be taken as a composite. On that approach it seems to me more in accordance with ordinary legal usage to read the reference to a "decision" as being to the totality of the decision, comprising both the formal decision and the reasons. It is true that that would render the reference to the "record of the decision" redundant, but if we are not taking a strict approach it would be enough to recognise that that sometimes happens.
32. My reading, on either approach, of the effect of the language used is reinforced by considering the purpose of a slip rule of this kind. It is not only in the case of a formal decision that it may be desirable for a court or tribunal to be able to correct accidental errors. Most obviously, such an error in its reasons might lead to further error, or in any event difficulty, on an appeal or in further proceedings in the same litigation, or in other cases where the decision was followed. The present case would, if the Tribunal's figure could be established to be an accidental error, be a good example of that; and clearly the applicant in Hazeltine thought that the missing "not" might create difficulties in understanding Whitford J's reasoning on the pending appeal. There may be other circumstances in which the impossibility of correcting an error could lead to other kinds of injustice or undesirable consequences. I can see no sufficient reason why the draftsman should have intended to frame a slip rule in terms that allowed correction in the one situation but not the other.
33. I note in this connection that rule 31 of the Tribunal Procedure (First-tier Tribunal) (Immigration and Asylum Chamber) Rules 2014 permits the correction of "any … accidental slip or omission in a decision, direction or any document produced by it". Whatever "decision" might cover if it stood alone, the italicised words appear to put it beyond doubt that the power extends to the written reasons for any decision. The TPC thus regarded it as desirable in that context that the slip rule should not be limited to formal decisions. It would be odd, to put it no higher, if we were to give rule 42 a construction which produced a different result in the Upper Tribunal.
34. Of course adopting such a construction means that rule 42 would be wider in its scope than rule 40.12 (1) of the Civil Procedure Rules, as construed in A (a Child). That may be untidy, but I do not regard it as unacceptable. The reasoning in A (a Child) depended on the established meaning of the phrase "judgment or order" and not on any view taken by the Court that a wider construction would be undesirable on grounds of policy or principle. Nor is there any reason to suppose that the drafters of the CPR themselves made a deliberate choice in favour of a narrow construction. It appears that they simply adopted the familiar formulation used in the RSC.^[4] Indeed if Hazeltine is correctly decided, any inconveniences caused by a narrow construction of the slip rule in the CPR itself are removed by the fact that there is an inherent jurisdiction to correct the reasons for a judgment or order.
35. I would for those reasons hold that the Upper Tribunal has power under rule 42 to correct clerical mistakes and other accidental slips or omissions not only in its formal decisions but in its reasons for those decisions.
36. That being so, it is not necessary that I reach a view on Mr Singh's alternative submission that the Upper Tribunal has such a power under its inherent jurisdiction. I have, however, read Newey LJ's judgment, and I am strongly attracted to his view that Hazeltine was correctly decided; and if the High Court has the necessary inherent jurisdiction, there is no good reason why the Upper Tribunal should not do so also. There is of course no problem, as such, about a statutory tribunal enjoying some powers by way of "inherent jurisdiction" (as regards the Upper Tribunal see, most recently, BPP Holdings v Her Majesty's Revenue and Customs [2016] EWCA Civ 121, [2016] 1 WLR 1915, per Ryder LJ at para. 25 (p. 1923)); and in Akewushola v Secretary of State for the Home Department [1999] EWCA Civ 2099, [2000] 1 WLR 2295, Sedley LJ, albeit in passing, expressly acknowledged that such a tribunal would have an inherent power to correct "slips" in its decisions (p. 2301 D-E). My only hesitation is about whether the fact that, once it has made its final decision, a court or tribunal is functus officio means that it can make no corrections to its reasons except to the extent that its powers are preserved by an express slip rule. If the matter were free from authority I would hold that the post-promulgation correction of accidental errors in the reasons of a court or tribunal did not involve the exercise of any its powers in a way that would attract the operation of the functus officio rule: that is also Newey LJ's view (see para. 50 below). However, there are observations by both Patten LJ and Black LJ in A (a Child) – see their judgments at paras. 24 and 41 (pp. 4459 and 4463) respectively – which suggest otherwise and are at least arguably ratio^[5]; and in Space Airconditioning Plc v Guy [2012] EWCA Civ 1664, [2013] 1 WLR 1293, Mummery LJ appeared to assume that the power to correct a judgment (in the broader sense) was lost at the moment that the formal judgment or order was entered – see para. 53 of his judgment (p. 1303H). We were not addressed in any detail on this aspect, and since the point is not determinative I prefer not to express a concluded view.
37. I do not anticipate any ill-consequences from a decision that the Upper Tribunal has a power of this kind; and indeed, as noted above, it appears that the First-tier Tribunal already does. The Tribunal will not, of course, be obliged to correct every slip in its written reasons that may subsequently come, or be brought, to its attention. Many decisions, however carefully written, contain minor errors of one kind or another, but in the great majority of cases there is no purpose in formally correcting them. It is only where there is some real value in doing so that the exercise may need to be undertaken.
38. I would add, finally on this aspect, that slips of the kind with which we are concerned here would in the great majority of cases be picked up and corrected without any need to invoke the slip rule if the decision and reasons were circulated to the parties in draft prior to the making of the formal decision; and it is a great pity that that course was not taken in this case. It must be for the Upper Tribunal itself to decide whether as a matter of practice all reserved decisions should be pre-circulated, and there may be reasons why that might not be a good idea; but the President may wish to consider making the practice mandatory at least in cases of particular substance.

(B) APPLICATION IN THE PRESENT CASE

39. Now that it is established that the only route by which the Secretary of State seeks to have the error in the Upper Tribunal's decision corrected is by the application of the slip rule – rather than by the exercise of a version of the English/Barke jurisdiction – there is a question whether this Court should have anything to say about whether it should be applied in the present case. The purpose of the slip rule is to allow the court or tribunal to correct errors of expression so as to express what it actually meant to say. Only it can say whether a particular error is indeed an error of expression and, if so, how it should be corrected.^[6] It follows that there would normally be no role for this Court. As regards the present case Mr Singh made it clear that the Secretary of State was prepared himself to apply to the Upper Tribunal to exercise its jurisdiction under rule 42, though he contemplated that we might wish to lend our weight to any such application. He fully acknowledged that only the Tribunal itself could say whether the error in paras. 106 and 196 was an error of expression – that is, that in its actual reasoning it had relied on the correct figure – or was substantive; and he also accepted that if, in view of the passage of time, the Tribunal was no longer sure which it was it should decline to correct it. He submitted that the Tribunal could and should be trusted to correct the error only if it was confident that it had indeed been no more than an error of expression.
40. I am sure that that is the right approach in the generality of cases. However, Ms Naik submitted that in the particular circumstances of this case it would not be proper for the Tribunal to exercise its jurisdiction even if it felt able to say with confidence that the reference to 0.01% was an error of expression; and that if that was so it would be appropriate for this Court to say so now rather than have any dispute deferred.
41. I am persuaded that if we are indeed satisfied that this could in no circumstances be a proper case for the application of the slip rule it is better that we should say so at this stage. If we do, I cannot think that the Secretary of State would take the matter further. I suppose the Tribunal could consider exercising its powers of its own motion, and if it did I do not believe that it would be formally bound by our views, but they would at the very least inform its own decision. In any event I think there is value in our considering Ms Naik's submission.
42. I can summarise Ms Naik's submissions as follows:

(1) She did not accept that it was self-evident that the Tribunal must have based its substantive reasoning on the (correct) figures in Mr Singh's speaking-note, and thus must have mis-expressed them in its decision. The substitution of "less than" for "about" raised the possibility that it had done its own calculations. Even if that were not so, the error had clearly crept in at some point between submissions and final decision, and it was just as plausible that it had done so before the Tribunal entered on its substantive decision-making as after: the process took many months. The description of the risk as "tiny" suggested that it was considering the 0.01% figure that it gave rather than the 0.1% figure given to it by Mr Singh.

(2) Given that there was thus a genuine question to be answered about what its thinking had been, it would be likely to be difficult for the Tribunal, coming to the question a year or more after the decision-making process, to say with confidence whether its error was substantive or only one of expression; and the difficulty was compounded by the fact that the question needed to be answered in the case of two judges.

(3) It was not a sufficient answer to say that if the error was substantive, or if the Tribunal was unable to say either way, it could be relied on to say so and to decline to make the correction sought. She disavowed of course any suggestion that the Tribunal would consciously misrepresent its recollection, but there was inevitably a risk of subconscious bias in favour of the less serious form of error. In any event, what mattered was not whether there was in fact any such risk but whether the fair-minded observer might believe that there was. She referred, by way of analogy, to the warnings of this Court in Barke that the power to ask for supplementary reasons should not be employed in cases where there is a real risk that any reasons given may, albeit usually subconsciously, be reconstructions of proper reasons rather than the unexpressed actual reasons for the decision made: see per Dyson LJ, giving the judgment of the Court, at paras. 44-46 (pp. 1389-90). An illustration of a request for clarification having been refused on that basis was to be found in Aerospace Publishing Ltd v Thames Water Utilities Ltd [2006] EWCA Civ 717: see per Rix LJ at paras. 12-13.

(4) She emphasised that the errors with which we are concerned go to the heart of one of the central issues in a country guidance decision of great importance to many refugees and other migrants facing return to Afghanistan. It would be unacceptable to leave in place a decision which was tainted by a reasonable perception of unfairness.

43. I have come to the conclusion, I have to say reluctantly, that those submissions are right. This is not a case of a straightforward error where it is clear what has gone wrong and how it should be corrected. Even if the additional zero in the percentage could safely be taken out, which is itself debatable, can the "less than" also be simply struck through ? The border-line between substantive error and error of expression may in this case be hard to draw. In these circumstances I do not think that it would be possible to exclude a real perception of unfairness if the Tribunal were simply to produce a corrected version of the decision that faithfully reflected the figures that it had been given by Mr Singh.
44. I have wondered whether it would be open to the Tribunal to seek to remove any perception of unfairness by offering an explanation of how the error(s) came about. It might in principle be the case that it would be able to refer to (even if it need not produce) documents – say, an e-mail exchange between the judges, or an earlier draft of the judgment – which demonstrated incontrovertibly that the decision was taken on the basis of the correct figures. It would be a shame, to put it no higher, if the appeal were to proceed on the basis of a version of the Tribunal's decision which could be shown, if the opportunity were given, not to represent its actual reasoning. Could it not be given guidance that it should make the corrections sought if but only if it was able to accompany them with a sufficiently circumstantial explanation to demonstrate that the errors were of expression only ? But the question only has to be posed to see that the cure is worse than the disease. A court or tribunal must state the reasons by which it justifies its decision; but the internal processes by which it formulates those reasons and arrives at its decision must remain a closed book.
45. I would add for completeness that it was part of Ms Naik's submissions that it would be inappropriate for the Tribunal to apply the slip rule simply because there was "an element of dispute" about the correction proposed. She took that phrase from the judgment of the Upper Tribunal (UTJ Bishopp) in Tager v Her Majesty's Commissioners of Revenue and Customs [2015] UKUT 663 (TCC): see at para. 21. It is entirely apt in the context in which it is used there, but I would not accept, if Ms Naik was so contending, that the fact that the parties do not agree about whether there is an error appropriate for correction under the rule is as such a reason for the Tribunal not making it. Given the nature and purpose of the jurisdiction the Tribunal alone knows whether it made a slip in expressing what it intended to say, and the parties' views on that question are immaterial. As the Upper Tribunal (Mark Ockelton V-P and UTJ Martin) put it at para. 9 of its decision in Katsonga [2016] UKUT 00228 (IAC):

"It is because the judge can use the slip rule only to make his original meaning plain rather than to change his original decision, that the Civil Procedure Rules and the Tribunal's Procedural Rules contain no provision for consultation with the parties. Indeed it is difficult to see that the parties ought to have any input into the judge's expression of what he originally meant."^[7]

46. In my view, therefore, in the very particular, and unusual, combination of circumstances obtaining in this case it would be wrong for the Upper Tribunal to use its powers under rule 42 to correct the errors in question even if it were satisfied that they are errors of expression. It has made its decision and the right course is for the appeal to proceed on the basis of its reasons as expressed.

DISPOSAL

47. There was no application before us, and for the reasons explored above I do not believe that we need or should make any formal ruling. The substance of our decision, however, if My Lord and My Lady agree, is that the Upper Tribunal would have jurisdiction to correct the error in its decision if it was a mere error of expression but that it would not be right for it to exercise that jurisdiction in the particular circumstances of this case.

Lord Justice Newey:

48. I agree that, for the reasons given by Underhill LJ, rule 42 is wide enough to allow the Upper Tribunal to correct clerical mistakes and other accidental slips or omissions in its reasons as well as its formal decisions. I agree with him, too, that it would not be appropriate for the Upper Tribunal to exercise the jurisdiction in the specific circumstances of the present case. The only issue on which I wish to add anything is whether Hazeltine Corporation v International Computers Ltd was correctly decided. Our interpretation of rule 42 means that the point can be of little importance in the context of the Upper Tribunal, but it retains significance for Court proceedings, particularly in the light of the decision in In re A (a Child).
49. No one doubts that a judge asked to approve a transcript of an oral judgment can make substantial revisions to it. There is no question of his being tied to the precise words that he used at the time. He can alter the wording, and also the structure, so that the judgment reflects his intentions. He may even insert a finding that he did not originally mention. In Secretary of State for Trade and Industry v Rogers [1996] 1 WLR 1569, which involved an application for an order under the Company Directors Disqualification Act 1986, Sir Richard Scott V-C, with whom Roch and Henry LJJ agreed, said (at 1578):

"Nor, in my judgment, can it be said that a judge has acted improperly in adding to the approved transcript a finding that he had not mentioned in the judgment as delivered extempore. In giving any judgment, extempore or reserved, a judge ought to mention the important features of the case that have led to the conclusion he has reached and the decision he has made. If, in a director's disqualification case, a judge concludes, whether rightly or wrongly is for this purpose immaterial, that on the facts of the case the director has acted dishonestly and that for that reason, perhaps among other reasons, a disqualification order for a particular period ought to be made, it seems to me plain that the judge should say so….

Harman J.'s amendment to the transcript of his extempore judgment made it clear that his conclusion that Mr. Rogers was unfit and that a disqualification period of eight years was appropriate was based in important part on a finding that Mr. Rogers's conduct had been dishonest. That being so the judge was right to make the amendment to the transcript and cannot be criticised for having done so."

50. By the time a judge is presented with a transcript for approval, the relevant order will almost inevitably have been perfected. That, however, is immaterial. It cannot be suggested that a judge's ability to revise a transcript of an oral judgment comes to an end, or becomes more limited, at the moment an order is entered. He is not, for this purpose, functus officio.
51. The extent to which a judge can properly change a judgment that was given in writing is plainly far more limited. In practice, however, minor corrections to such judgments are sometimes made even though the orders have already been drawn. It is not that unusual for a judgment put on BAILII to be replaced by a revised version, and I should be surprised if that happened only where the order in question had not yet been entered. Again, where a judgment is to be included in the official law reports, the All England Law Reports, or certain other series, the draft report will be sent to the judge for approval and an eagle-eyed law reporter might query, say, whether a cross-reference to "paragraph 25 above" should in fact have been to "paragraph 26 above" or whether, where the handed-down version of the judgment said "section 6", the judge had really meant "section 8". As I understand it, a judge can properly accede to such a suggestion, regardless of the fact that the order may have been perfected long before. I referred to what I take to be an example of this in Briggs v Gleeds [2014] EWHC 1178 (Ch), [2015] Ch 212. As I mentioned in paragraphs 116-117, Millett J appears to have excised the words "and 'accrued'" when approving the report of In re Courage Group's Pension Schemes [1987] 1 All ER 528 for the All England Law Reports.
52. In Powell v McFarlane (1977) 38 P&CR 452, Slade J said, according to the report, that "factual possession" "must be a single and conclusive possession" (see 470). Endorsing that statement of the law in JA Pye (Oxford) Ltd v Graham [2003] 1 AC 419, Lord Browne-Wilkinson quoted Slade J (at paragraph 41) as having said that "factual possession" "must be a single and [exclusive] possession". Lord Browne-Wilkinson will have inserted the square brackets on the basis that Slade J meant to say "exclusive" rather than "conclusive". Had Slade J been provided with a draft of the report and spotted the error, it seems to me that he could properly have made the change himself. The functus officio principle would not have applied either because his judicial role would have extended to making the correction or because mere correction of an error of expression is not to be regarded as involving an exercise of judicial power to which the principle would attach.
53. Such matters appear to me to point towards the legitimacy of the approach that Whitford J took in Hazeltine. As, moreover, Underhill LJ has explained in his judgment, it is desirable that Courts should have the power to correct the reasons for a judgment as well as the "end product". In Green v Adams [2017] EWFC 52, [2017] 4 WLR 140, Mostyn J remarked (at paragraph 17) that "the power to correct a judicial mistake … obviously exists", citing Hazeltine. I agree that there is such a power. In my view, Whitford J was right to consider that the Court has an inherent jurisdiction to correct slips in judgments even when the relevant orders have been entered. The power must, however, be a very limited one, to be exercised sparingly.

Lady Justice Nicola Davies:

54. I have read the judgment of Underhill LJ and for the reasons given I agree that it would be wrong for the Upper Tribunal to use the powers under rule 42 to correct the identified errors. However, I do not share the measure of reluctance expressed at [43] in concluding that the submissions of the Appellant identified at [42(1)-(4)] were well founded. The two errors relied upon by the appellant were not straightforward. To clarify the same, in particular the error resulting from the substitution of "less than" for "about" would involve or be perceived to involve a measure of reasoning by the Upper Tribunal. Such a process could or would result in a real perception of unfairness.
55. I have also read the judgment of Newey LJ. Like Underhill LJ, I am attracted to Newey LJ's view that Hazeltine was correctly decided but I also share the hesitation expressed by Underhill LJ at [36] in respect of the operation of the functus officio rule. As this aspect of the argument was not addressed in detail, for the purpose of this judgment I do not express a concluded view.

Note 1    As regards the latter alternative, if the point had been picked up at the draft stage the Tribunal could simply have deferred formal promulgation and re-considered. If it had only been picked up at the permission stage, it would probably have been too late for it to reconsider: rule 45 of the Upper Tribunal Rules 2008 does give the Tribunal a power to review its decision in the response to errors identified on an application for permission to appeal, but the grounds are very limited and would not appear to apply to the circumstances of this case. But it could have considered exercising its power under the slip rule, albeit that there would have been difficulties about that course, as discussed below.    [Back]

Note 2    I should say for completeness that we heard no submissions on the legislative history of the phrase “a decision or record of a decision”, but the first use that I have been able to find is in section 6 of the National Insurance Act 1974, which is the slip rule covering decisions taken under the Social Security Acts. I suspect that the language reflects the way in which such decisions were framed. But that history cannot cast useful light on its meaning in the 2007 Act, where, as I say, the draftsman must be taken to have appreciated that it would have to apply to a wide variety of different kinds of decision by different Tribunals and Chambers.     [Back]

Note 3    That is what happens in practice, with at least some classes of decision. In fact one of the Upper Tribunal decisions in our bundle of our authorities (Katsonga) is headed simply “Decision and Reasons” and is an undifferentiated piece of text, with no separate decision notice.    [Back]

Note 4    There is no way now of establishing why in its original form, which goes back to the nineteenth century, the slip rule referred only to a “judgment or order”. But it is worth bearing in mind that until comparatively recently judgments (in the wider sense) giving the reasons for a judgment (in the narrower sense) were not delivered in written form. All judgments in the High Court and Court of Appeal were delivered orally, even if a transcript might be made and the judgment might in some cases in due course be reported. In those circumstances it is perhaps unsurprising that no need was recognised for a power to correct anything but the formal decision.    [Back]

Note 5    In my own judgment I do refer to Hazeltine and at least contemplate that it may be correctly decided (see para. 34 (p. 4462F)); but it is not referred to by Patten LJ or Black LJ.    [Back]

Note 6    That does not mean that the jurisdiction can only be exercised by the judge who made the decision in question: in principle the power is vested in the court or tribunal – see Cripps (above), per Sir John Donaldson MR at p. 695 B-C. But in cases where the nature of the error and how it should be corrected are not self-evident, it may be that another judge would not be able properly to exercise the jurisdiction.    [Back]

Note 7    My citation of Katsonga should not be taken as implying approval of the proposition in the judicially-drafted headnote that “the ‘Slip Rule’… cannot be used to reverse the effect of a decision”, which if taken out of context may be misleading. If, say, a “not” were accidentally omitted from a declaration or injunction its correction might well reverse what would otherwise be the effect of the decision, but it is hard to see why it should for that reason be illegitimate: indeed it might be thought to be the paradigm of the kind of case for which the slip rule was required.     [Back]