WM Morrison Supermarkets Ltd v Revenue and Customs (VAT - cool-down rotisserie chickens - whether "hot food")

In CCE v National Westminster Bank plc , [2003] STC 1072 (" NatWest "), the taxpayer complained of unfair treatment (not being treated in the same way as other comparable car leasing companies) in the context of an appeal under section 83(t) VATA and section 84(10) VATA against a decision of the commissioners not to repay overpaid VAT. Jacob J held that the tribunal did not have jurisdiction to determine the unfair treatment point. Section 83(t) provides that "... an appeal shall lie to a tribunal with respect to any of the following matters ... (t) a claim for the repayment of an amount under section 80". HMRC had argued that the complaint about unfair treatment was essentially a complaint about their conduct and the proper remedy for unfair treatment is judicial review. Referring to Corbitt , Judge J said (at [49]) that "There is authority which supports the conclusion that general conduct towards taxpayers is outwith the Tribunal's jurisdiction". Moses J had reached the same conclusion in Marks & Spencer plc v CCE , [1999] STC 205 (" Marks & Spencer "), when he observed (at p246):

In Oxfam v Revenue and Customs Commissioners, [2010] STC 686 (" Oxfam "), the principal issue was the extent to which Oxfam was able to recover input tax in accordance with an agreed method for apportioning business and non-business expenditure. The subsidiary issues were whether the FTT could enforce an agreement between the taxpayer and HMRC, and whether it could give effect to a taxpayer's legitimate expectation. Those issues were of academic interest only as it was decided that, as a matter of fact, there was neither a binding contract nor a legitimate expectation, but the judge went on nevertheless to consider the extent of the tribunal's jurisdiction.

At [62] Sales J identified the legislative provision which conferred jurisdiction on the Tribunal, in that case section 83 of VATA:

Then, at [63], he said:

After recording that the parties had agreed that the tribunal's jurisdiction extended to the contract issue, and saying that he concurred in that view, he went on to say:

The judge then went on to consider whether there was any authority which compelled him to a different conclusion (mentioning, among others, Corbitt , NatWest and Marks & Spencer ), and decided that there was not.

In HMRC v Hok Limited , [2012] UKUT 363 (TCC) (" Hok "), Warren J and Judge Colin Bishopp (the then Presidents of the Tax and Chancery Chamber of the Upper Tribunal and the Tax Chamber of the First-tier Tribunal) dealt with an appeal by the taxpayer against penalties imposed on it by HMRC for failing to submit an employer's end of year return by the due date for its submission. Although it was conceded that the return was late and that there was no reasonable excuse for the lateness, the FTT had concluded, on grounds of fairness, that the taxpayer should be penalised for only the first of the five months which passed before the return was submitted. HMRC appealed against that decision on the basis that the FTT had no jurisdiction to discharge such penalties if they are properly due. Section 100 of the Taxes Management Act 1970 provided that:

Relying on Corbitt , NatWest and Marks & Spencer , the UT held (at [41]) that, "There is in our judgment no room for doubt that the First-tier Tribunal does not have any judicial review jurisdiction". At [43] they explained that the legislation creating the current tribunal system pointed to this being the only conclusion. They said:

At [49], having set out Sales J's observations on the jurisdiction issue in Oxfam , they observed that:

Having reviewed cases said to lead to a different conclusion and decided that they did not support the idea that the tribunal has jurisdiction to apply general principles of public law, they made these comments on Oxfam :

From that sentence it becomes clear that the issue in that case and the issue here are quite different. There, the tribunal was required to decide the amount of input tax which Oxfam could recover, a question which, as Sales J said at [63], comes four-square within the ambit of s 83(1)(c) of VATA. Here, the question is not the amount of a penalty, or even whether one is due as a matter of law—there is no dispute that s 98A was engaged, and that it imposed a liability for five monthly penalties of £100 each—but whether HMRC should be precluded from imposing the penalties prescribed by that section, or from collecting them if imposed. That, in our judgment, is a quite separate question of administration, one which, in accordance with the authorities to which we have already referred, is capable of determination only by way of judicial review and therefore not by the First-tier Tribunal."

In HMRC v Noor [2013] UKUT 71 (TCC) (" Noor "), a UT of the same composition as that in Hok considered whether the FTT had any jurisdiction to consider a legitimate expectation ground on an appeal against a decision concerning a claim for input tax credit under section 83(1)(c) VATA. The UT held that it did not. Although the UT acknowledged that it was dealing with an appeal under section 83(1)(c), the UT was very clear from the outset that its observations would be of wider impact; see its comments in [1].

At [30], having built out from its analysis in Hok , the UT concluded:

Nevertheless, at [31] they observed:

Turning to Oxfam , the first question was the status of Sales J's remarks. In Hok , the UT had observed that his comments on jurisdiction were obiter, but now they changed their minds. Having reviewed the position in great detail, they concluded (at [50]) that Sales J's "conclusion was a matter of decision. In so doing, we depart from our own (clearly obiter) remark to the contrary in Hok. " Even though the UT concluded that Sales J's comments on jurisdiction in Oxfam were a matter of decision, they considered that they were not bound by that decision, nor did they consider that, although not bound by his decision, they ought to follow it; see [94].

In reaching their decision, the UT explained (at [77]) why they did not consider that it was plausible that Parliament would have intended to give the FTT a wide power of judicial review in relation to all matters within section 83(1), which they thought (see [75]) would be the result of Sales J's comments, on the basis that "His approach to the 'ordinary and natural' meaning of s 83 applies to all its paragraphs; there is no hint in his reasoning that it turned somehow on the particular wording of para (c)." Most important here was the structure of the tribunal system created by TCEA 2007, which led the Tribunal to conclude (at [77](c)) that "It is simply inconceivable that Parliament would have contemplated conferring a similar power [to the judicial review powers of the UT] on the FTT".

They considered that section 83(1)(c) was concerned only with "the right to a credit arising under the terms of the VAT legislation (including, on one view, HMRC's care and management powers)". The words 'with respect to' in the opening language of section 83(1) were not "wide enough 'to cover any legal question capable of being determinative of the issue of the amount of input tax which should be attributed to a taxpayer' at least not in relation to the 'amount of input tax' which should be attributed to a taxpayer. ... financial adjustment within the phrase 'amount of input tax'. On that basis, Sales J's reading goes too far, in our view. It departs from the natural meaning of s 83(1)(c) which, reading the subsection as a whole, is focused on the large number of decisions on rights and obligations under the VAT legislation which HMRC have to make and in respect of which a specialist tribunal is provided."; see [92]-[93].

Similar conclusions, that the terms of TCEA make it clear that the FTT should not have a judicial review function, were reached in the context of very different pieces of tax legislation by the UT (in Reed Employment plc v HMRC , [2014] UKUT 160 (TCC) (" Reed "), and the Court of Appeal (in Trustees of BT Pension Scheme v HMRC , [2015] EWCA Civ 713 (" BT Pension Scheme ")). Although dealing with section 84(10) VATA, the Court of Appeal in Metropolitan International Schools Ltd v HMRC , [2019] EWCA Civ 156 (" MIS "), was of a similar view. Newey LJ (with whom David Richards and McCombe LJJ agreed) said this:

That would be a very surprising result. ...

Mr Ramsden did not attempt to persuade us that the UT was wrong in Noor . Were, however, his contentions as to the ambit of s 84(10) of the VATA well-founded, it would seem that the FTT had, after all, a wide jurisdiction to rule on public law issues and, in particular, legitimate expectation claims. The jurisdiction would, moreover, have been conferred through a provision introduced in response to the Corbitt decision (viz s 84(10)) ('by the back door', as Miss Mitrophanous would say), rather than under s 83, the main appeals section. Further, legitimate expectation (and, seemingly, other public law) arguments could be raised in the FTT without any need to satisfy the requirements as to obtaining permission and time limits that govern applications for judicial review (see CPR 54.4 and 54.5). It is highly improbable that Parliament intended this when it enacted what has now become s 84(10)."

Beadle v HMRC , [2020] EWCA Civ 562 (" Beadle "), concerned a penalty notice for non-payment of a partner payment notice ('PPN' - the PPN regime was introduced by Finance Act 2014 as part of a set of provisions enabling HMRC to require accelerated payments to be made by taxpayers on account of disputed or unresolved tax liabilities in certain prescribed circumstances) issued to the taxpayer in relation to certain film scheme arrangements the taxpayer had participated in, requiring the taxpayer to make a payment. The PPN regime confers no statutory right of appeal to a specialist tribunal against a PPN, but a partner who receives a PPN is entitled to make written representations to HMRC objecting to the PPN.. The taxpayer made representations to HMRC challenging the validity of the PPN, but those were rejected. The taxpayer did not pay the amount due under the PPN in time and was issued with a penalty notice. He appealed the penalty notice, asserting, inter alia, that the PPN was a nullity in law and/or the amount payable under the PPN should have been zero so that the penalty should also have been zero. The taxpayer contended (inter alia) that, in the absence of any statutory appeal process against PPNs, collateral challenges to PPNs on public law grounds could be made by a taxpayer in the course of a statutory appeal against a penalty notice for non-compliance with a PPN

In her judgment, in which she ultimately concluded that the tribunal did not have jurisdiction to determine the validity of the underlying PPN in the penalty appeal proceedings, Simler LJ started with the general rule (derived from O'Reilly v Mackman , [1983] 2 AC 237 at 285) that it is an abuse of the process of the court to permit a person to challenge a public law decision by means other than judicial review, but acknowledged (at [44]) that this rule is:

She went on to hold (at [45], disagreeing with Mr Beadle's counsel) that, whilst a statute might expressly exclude a public law challenge, "the express words used by a statutory scheme looked at in isolation may not be sufficient on their own to restrict or exclude public law challenges, but that may be the clear and necessary implication when the relevant statutory scheme is construed as a whole and in light of its context and purpose". As to the correct approach to take to construing the relevant statutory scheme, she said this (at [47]):

Looking at the statutory scheme in that case, she concluded (at [48]) that it is a clear and necessary implication of the PPN regime, construed as a whole and in light of its statutory purpose, that the ability to raise a collateral public law challenge to the validity of the underlying PPN is excluded at the penalty and enforcement stages.

In Zeman the taxpayer appealed to the FTT against an assessment to VAT on the grounds that it had a legitimate expectation that it would not be assessed to VAT on certain supplies. It said that its legitimate expectation arose from statements made by HMRC. The FTT dismissed its appeal, finding that KSM could not rely on the principle of legitimate expectation because it had not acted reasonably in relying on HMRC's statements. There were two issues in the taxpayer's appeal to the UT:

The UT held that the FTT did not err in its decision that the taxpayer did not have a legitimate expectation that it would not be taxable on the supplies in question. It then observed (at [20]):

The assessment in that case was raised under section 73(1) VATA and the appeal was brought under section 83(1)(p) VATA ("an appeal ... with respect to ... an assessment").

The UT began by observing (at [27]) that "the nature of the FTT's jurisdiction depends on the proper construction, in the context of the statutory provisions to which it relates, of the statutory provision by which it is given, in this case, s 83(1)(p)". They described that as the beginning rather than the end of the enquiry. At [29] they acknowledged the exclusivity principle derived from O'Reilly v Mackman , but considered it to be well established that the principle is subject to exceptions, one of which was "certain cases where a public law defence is raised in a private law action" and here they referred to Wandsworth London BC v Winder, [1985] AC 461 and Pawlowski (Collector of Taxes) v Dunnington [1999] STC 550 . At [30] they said that, "The question in this case is how the exclusivity principle operates in the context of a statutory scheme which contemplates private enforcement action being taken against a defendant." And that "An important part of that analysis is to determine the proper approach to be taken to construction of the relevant statutory language."

At this point the UT referred to the CA decision in Beadle , citing paragraph [44] of Simler LJ's judgment and her comment that the exclusion need not arise expressly but might arise by clear and necessary implication when the relevant statutory scheme is construed as a whole and in light of its context and purpose. The UT then observed (at [34]):

The UT started their analysis of the position where section 83(1) VATA is concerned with Sales J's judgment in Oxfam . They considered (at [42]) that Sales J in Oxfam had held that, depending on the nature of the issues falling within the scope of a particular sub-heading or subsection of section 83, it may well be that public law principles do fall within the scope of the appeal jurisdiction that subsection confers. They acknowledged that the UT took exactly the opposite view of the same issue under s 83(1)(c) in Noor, where it had disagreed with Sales J's view that as a matter of ordinary language in context the words 'with respect to' were wide enough to cover any legal question relevant to the issue of the amount of input tax attributable to the taxpayer, on the basis that any result of giving effect to the legitimate expectation would not affect the 'amount of input tax'. They noted that this approach (which they described at [46] as "draw[ing] a distinction between determining of the amount of tax due (which falls within the appeal jurisdiction), and other matters (which do not)") echoes other decisions such as NatWest and Aspin .

At [48] the UT observed that cases are likely to depend on the statutory language in question. So, Aspin could be explained by the limitation in section 50 TMA on the actions the General Commissioners could take, and NatWest by Jacob J's reading of s 83(1)(t) VATA, that it conferred an appeal jurisdiction only where the challenge was that an amount of VAT was not in fact due. It did not confer jurisdiction in a case where the relevant VAT amount was due but was said to be repayable for an extraneous reason.

As regards the provision here, the UT noted that section 73(1) VATA is permissive; it enables but does not require HMRC to raise an assessment. Also, section 83(1)(p) VATA ("an appeal ... with respect to ... an assessment") is wider than section 83(1)(c) (an appeal with respect to the amount of an assessment). It analysed the leading cases on best judgment ( Rahman (t/a Khayam Restaurant) v Customs and Excise Comrs (No 2) [2002] EWCA Civ 1881 , and Customs and Excise Comrs v Pegasus Birds Ltd [2004] EWCA Civ 1015 ) and concluded that they did not lead to a different conclusion, not least because "neither decision was concerned with defining definitively the full scope of the appellate jurisdiction under s 83(1)(p). They were concerned with defining the scope of the 'best of judgment' test and with the consequences of breach of that test". Indeed, the UT observed that:

Assume for example a case in which the taxpayer's defence is that an assessment was made dishonestly or maliciously in knowing disregard of an undertaking not to assess. HMRC's argument would be that that defence has no place on an appeal under s 83(1)(p), because it relates to the decision to assess. It is true that it does, but that is not the same as saying that it relates only to that question. On the contrary, it seems to us it might equally well be said to be relevant to the process assessment, because it is difficult to see how an assessment made in knowing disregard of such an undertaking - whether binding in contract or under general principles of public law - could be said to be an assessment made to best judgment. ...

In such circumstances, it seems to us there are good policy reasons for not adopting a construction of s 83(1)(p) which strictly limits the appellate jurisdiction of the FTT in the manner identified in the Gore decision at [30] (see [65] above), and which therefore excludes consideration of a legitimate expectation argument. We refer again to the comments of Sales J in Oxfam quoted at [39] above. Were one to adopt such a restrictive approach, there would be an obvious risk of duplication, delay and potential injustice given the potential for disputes to arise as to which forum any particular challenge should be brought it."

The UT concluded (at [84]):

The Executors of David Harrison (Deceased) v HMRC , [2021] UKUT 273 (TCC) , involved a challenge to HMRC's decision not to accept a notice electing for fixed protection for a pension scheme. There was a deadline for submitting a notice electing for fixed protection but HMRC had a discretion to accept a notice served after that date. One issue was whether the FTT had jurisdiction to review the exercise of HMRC's discretion. Relying on paragraph 45 of Simler LJ's judgment in Beadle , the appellants in that case argued that the relevant regulations would need to contain express words or a necessary implication to exclude their right to make a public law challenge to HMRC's exercise of discretion. The UT rejected that argument commenting:

Paragraphs 44 and 45 of Simler LJ's judgment in Beadle cannot be read as laying down some general proposition to the effect that the FTT always has jurisdiction to consider public law challenges to HMRC decisions unless express words or "necessary implication" exclude that jurisdiction. Her conclusion was more limited and addressed at the specific situation of persons defending themselves against enforcement action brought by a public body and seeking, in the context of their defence, to make a public law challenge to either the enforcement action itself or some antecedent action."

The appellants also argued that the UT decision in Zeman had extended the principle set out in paragraph 45 of Simler LJ's judgment in Beadle and urged the UT to adopt the same approach. The UT responded as follows (at[34]):

At [36] they rejected HMRC's submission that clear words would be needed in order for the Regulations to give the Appellants the right to challenge the exercise of HMRC's discretion before the FTT, saying:

Caerdav Limited v HMRC , [2023] UKUT 179 (TCC) (" Caredav "), concerned an appeal under section 83(1)(b) VATA (which permits appeals to the FTT with respect to "the VAT chargeable... on the importation of goods from a place outside the member States") against a C18 demand notice for VAT and customs duty. The FTT had held that, as a matter of statutory construction of section 83(1)(b), it did not have jurisdiction to consider legitimate expectation as a ground of appeal and it was only a matter that could be raised on judicial review.

At [152] the UT endorsed the starting point," that appeal grounds which concern public law arguments should be pursued in judicial review proceedings rather than before the FTT. However, we, like the FTT, accept that the FTT may have jurisdiction to consider appeal grounds based on public law arguments (such as legitimate expectation) depending on the statutory provisions under consideration."

At [154]-[155] the UT drew a distinction between the rights of appeal under section 83(1)(b) and that under section 83(1)(c), both of which it thought were simply concerned with whether the conditions prescribed for a charge to arise under the legislation are present and the amount of the charge. They contrasted this with the right of appeal in section 83(1)(p), discussed in Zeman , which (they said) "provides a right of appeal against the discretion of HMRC whether to make an assessment under section 73(1)". In their words (at [155]):

Ms Sloane's position is that the issue of jurisdiction has already been decided by the UT in Zeman . The ordinary course is for this Tribunal to follow it (irrespective of whether this Tribunal would have decided the issue in the same way or can see force in some arguments criticising its rationale). As an inferior tribunal, we can only depart from Zeman if we consider that the comments on jurisdiction were obiter or if the decision was reached per incuriam. Even if we do not consider that Zeman is binding on us, we should only depart from it if we consider the decision to be plainly wrong.

In Queenscourt Limited v HMRC , [2024] UKFTT 460 (TC) (" Queenscourt "), the FTT (Judge Robin Vos and Mrs Catherine Farquharson) was faced with a submission that, following Zeman , the FTT had jurisdiction to consider arguments based on legitimate expectation in the context of an appeal under section 83(1)(t) VATA. They concluded (at [142] that, because the UT had already decided that the appeal failed because Zeman did not have the legitimate expectation which it alleged and that (in its words) was sufficient to dispose of the appeal, the UT's comments on jurisdiction were obiter. However, being a decision of the Upper Tribunal and, having addressed the question of jurisdiction at some length, the FTT considered that the decision was potentially of significant persuasive authority. At [163] the FTT said that the approach taken in Zeman was correct in the light of the decision of the Court of Appeal in Beadle and (at [166]) "that the First-tier Tribunal (and for that matter any other Tribunal) is required to follow the approach set out in Beadle at [44]". They concluded (at [175] that there was nothing in the statutory scheme of s 80(4A) and s 83(1)(t) VATA which either expressly or implicitly excludes the Tribunal's jurisdiction to consider public law arguments and, in particular, arguments based on legitimate expectation, in relation to HMRC's decision to make the recovery assessments. They then said:

Indeed, it appears to us that the underlying assumption in Metropolitan International (a case dealing with s 84(10) VATA rather than s 83(1) VATA) appears to be that the First-tier Tribunal has no jurisdiction to consider arguments based on legitimate expectation in the context of s 83(1)(p) as it would otherwise have been unnecessary for the appellant to rely on s 84(10) VATA, although the Court of Appeal did not say this in clear terms.

However, as we have explained, we can find nothing in the statutory scheme which is relevant to this appeal which suggests that Parliament intended to exclude the jurisdiction of the Tribunal to entertain arguments based on public law grounds. We were perhaps hindered in that task by the fact that HMRC did not address the detailed analysis of the relevant legislative provisions in their submissions and it may well be that another Tribunal in the future comes to a different conclusion on this point with the benefit of fuller submissions."

In Treasures of Brazil Limited v HMRC ("Treasures of Brazil") , [2024] UKFTT 929 (TC) , the FTT (Judge Malcolm Frost and Mr Derek Robertson JP) needed to decide whether it had jurisdiction to consider a legitimate expectation argument on an appeal under section 83(1)(p). On the question whether Zeman was binding authority, the FTT said:

In Zeman , the Upper Tribunal first decided that the taxpayer did not in fact have a legitimate expectation, before considering the question of whether or not the First-tier Tribunal would have had jurisdiction to consider such issues.

The fact that the Upper Tribunal decided the question of whether or not the taxpayer had a legitimate expectation at an early stage in its reasoning meant it did not need to consider the jurisdiction question. However, despite not needing to, the UT did in fact go on to consider the jurisdiction point.

Therefore, the entire decision of the Upper Tribunal was (i) that the First-tier Tribunal had jurisdiction but (ii) that there was no legitimate expectation. If the Upper Tribunal had decided there was no jurisdiction then the legitimate expectation question would have itself been redundant. The jurisdiction point was therefore a constituent part of the decision made. We do not consider that the fact that the Upper Tribunal could have chosen not to determine the jurisdiction question means that it is open to this Tribunal to treat the jurisdiction question as obiter.

We are supported in that view by authorities such as Jacobs v LCC [1950] A.C. 361, (at p369 per Lord Symonds):

In any event, even if the Upper Tribunal's analysis in the Zeman case were not binding upon us, we consider that it is a correct statement of the law.

As such, we conclude this Tribunal does have jurisdiction to consider the Appellant's legitimate expectation argument."

A differently constituted FTT (Judge Natsai Manyarara and Mr John Agboola JP) in United Carpets (Franchisor) Ltd v HMRC [2025] UKFTT 895 (TC) (" United Carpets "), explicitly agreed (at [238]) with the propositions in paragraphs [41]-[45] of Treasures of Brazil .

Ms Sloane says that the approach taken in Treasures of Brazil effectively mirrors the reason why the UT in Noor reconsidered its decision in Hok (that Sales J's comments in Oxfam on jurisdiction were obiter) and held that Sales J had to decide that he had jurisdiction in order to go on and consider the legitimate expectation point, and so his comments on jurisdiction were a matter of decision.

As to whether Zeman was decided per incuriam, Ms Sloane says that HMRC overstate that principle. It does not apply just because a court did not consider all relevant authorities or might have reached a different conclusion. She referred us to the recent Court of Appeal decision in Merck Serono SA v The Comptroller-General of Patents, Designs and Trade Marks , [2025] EWCA Civ 45 (" Merck Serono "), where Lewison LJ (with whose judgment Birss and Arnold LJJ agreed) said this about the per incuriam concept:

They added:

Plainly there was no authority binding the court in Newron which would have compelled a different decision. The court in Newron was faced with a choice between Neurim and Santen both of which were decisions of the CJEU. Nor, in my judgment, was there an inconsistent statutory provision. The power to depart from retained EU law is not inconsistent with deciding what EU law was. In Miliangos v George Frank (Textiles) Ltd [1975] QB 487, 503 Lord Denning MR pointed out that "a case is not decided per incuriam because counsel have not cited all the relevant authorities or referred to this or that rule of court or statutory provision". Moreover, in Duke v Reliance Systems Ltd [1988] QB 108, 113 Lord Donaldson MR said:

HMRC criticise the UT in Zeman for not addressing MIS , but that concerned a different issue (section 84(10) VATA) and did not decide the basis of jurisdiction as canvassed in Zeman . There are no authorities not addressed in Zeman which would (as opposed to might) have led to a different analysis or outcome.

As to whether Zeman is plainly wrong on jurisdiction, which would (Ms Sloane accepts) be a reason not to follow it if it is not binding, Ms Sloane readily acknowledges that there are many arguments which can be made in support of a different analysis. It is no part of her case that Zeman is so compellingly correct that it is bound to be followed were the same jurisdictional issue to arise in the UT or a higher court. It is, however, very much her case that none of the questions that can be raised around Zeman establish that Zeman is "plainly wrong". This is an area of law which is "not straightforward" ( Noor at [82]) and the authorities "present a somewhat fragmented picture" ( Zeman at [36]).

HMRC argue that the UT failed to appreciate that Beadle applies only to enforcement proceedings in the sense of debt collection/recovery proceedings. That is not obviously right. Beadle itself was not an enforcement case, in the sense of debt collection/recovery proceedings by the Crown. It was an appeal against a penalty; the tax had been paid. If the Court of Appeal were confining the relevant principles to debt collection/recovery action against a taxpayer, it could have dealt with the case much more shortly by saying they did not apply to an appeal brought by a taxpayer against a penalty. HMRC argue that it may be "happenstance" whether a taxpayer makes a claim or faces an assessment. Beadle is itself an example of where the same situation can lead either to a repayment claim by the taxpayer or a claim for payment from HMRC. While each scenario has the same practical effect in terms of the taxpayer being out of the money, the mechanisms for legal proceedings and the applicable legal principles and remedies may differ fundamentally. There is nothing "plainly wrong" in that. In Zeman , the UT did carefully consider contrary authorities, in particular Aspin and Noor , and addressed them.

Mr Watkinson first submitted that Zeman is not binding on us because the UT's decision on jurisdiction was either obiter or reached per incuriam. He then went on to launch what might colloquially be described as an "all guns blazing" attack on that decision. At the end of his submission it was quite difficult to think of anything the UT said in Zeman on the jurisdiction point that he did not consider to be deeply suspect or just plain wrong.

Dealing first with the question whether what the UT said in Zeman about jurisdiction is a matter of decision or obiter, Mr Watkinson points to the decision of the Court of Appeal in R (oao Youngsam) v Parole Board , [2019] EWCA Civ 229 (" Youngsam "). He says that the ratio is the best or preferred justification for the conclusion reached: it is necessary in the sense that the justification for that conclusion would be, if not altogether lacking, then at any rate weaker if a different rule were adopted. In looking for the ratio decidendi of a case, the starting point is always the rulings and reasons given in the judgment(s) to justify the court's decision, read in the light of the facts of the case and the issues that arose. Generally, this is also where the inquiry ends. One can identify the ratio of Zeman with relative ease, because the UT did so itself. It is clear on the face of the UT's decision that what it said about jurisdiction was not "part of the best or preferred justification for the conclusion [it] reached" and that what it said about jurisdiction was obiter dicta. Whilst of course persuasive, it does not bind this Tribunal.

On the per incuriam point, Mr Watkinson says that, possibly because in Zeman the UT was making obiter observations, it did not deal with the earlier several authorities. It may also be that it did not have all the authorities put before it. These include Dollar Land, Marks & Spencer, Hok , Reed and BT Pension Scheme and MIS are not there.

Especially with the most recent authority, MIS not being before it, Zeman appears to be per incuriam.

Finally, the UT in Zeman does not appear to have considered the TCEA at all in its construction of the appeal provisions. As the authorities show, that is a serious error.

Mr Watkinson says that the remarks on jurisdiction in Zeman were wrong, per incuriam, and obiter as is any FTT decision purporting to apply that approach. In broad terms, his reasons for saying this are as follows:

The first question we need to answer is whether the comments on jurisdiction in Zeman are binding on us or are obiter. Here Ms Sloane points us to the way the UT in Hook and Noor analysed Sales J's comments on jurisdiction in Oxfam , and she says that we should approach Zeman in the same way that the UT ended up approaching Oxfam in Noor . In Hok , the UT had simply observed that they thought these comments were obiter. In Noor they said that they had changed their mind and inclined to the view that those comments were a matter of decision.

At [4] in Oxfam , Sales J observed that "as I explain below, I consider that Oxfam's claim based upon public law principles and the doctrine of legitimate expectation could properly have been raised in its appeal to the tribunal". He went on at [5] to say that he thought that the correct approach was to treat the legitimate expectation argument as a new argument raised on the appeal under VATA with the leave of the court and to rule upon it in the context of that appeal, applying principles of public law and not to grant permission for the same argument to be brought by way of judicial review. At [61] et seq he set his analysis of the Tribunal's jurisdiction, which links back to the words "as I explain below" in [4]. However, at [59] he had already expressed his conclusion "that there was no abuse of power on the part of HMRC in acting as they did. Therefore, Oxfam's claim based on legitimate expectation also falls to be dismissed."

At [43] in Noor the UT observed that Sales J could have " declined to decide which jurisdiction he was exercising, deciding simply that Oxfam's claim based on legitimate expectation failed whichever jurisdiction he was exercising" and (at [44]) they said:

However, another view (discussed by the UT at [47]) was that paragraph [5] in Sales J's judgment was a necessary part of his reasoning, so that his conclusion concerning the jurisdiction of the FTT was not obiter at all and the reasons for his conclusion which appear at [61] are to be read into para [4]. As a result his decision and reasoning in relation to jurisdiction inform everything which follows under the headings 'The Appeal against the Tribunal's Decision' and 'Legitimate Expectation' and the words of dismissal at [44] refer to the part of the VAT appeal that related to the VAT Tribunal's decision, not the (new) legitimate expectation argument. The words at the beginning of [44] in Oxfam ("The dismissal of the appeal is not the end of the matter ...") might be thought to support that view.

At [50] the UT in Noor said that it inclined to that view, but later (at [52]) observed "We acknowledge that the view that what Sales J said was simply obiter cannot be dismissed out of hand." Because of the strength of the view that what Sales J said was a matter of decision, the UT felt that they needed to decide whether they were bound by that decision (assuming Sales J's comments were a matter of decision) and concluded that they were not. As a result, it was not necessary for the UT to reach a conclusion about which view was correct since, on either view, it was able to make its own decision about the jurisdiction of the FTT.

Mr Watkinson referred us to Youngsam , where the Court of Appeal needed to decide what constituted the ratio of an earlier Supreme Court decision. Nicola Davies LJ (with whom Haddon-Cave LJ agreed) commented:

In my view the key words contained in the statement of Professor Cross are "treated by the judge as a necessary step in reaching his conclusion". "

Leggatt LJ went further, observing:

Even where a judge gives only one reason for a decision, there is generally no warrant for supposing that the ruling contained in the judgment was treated by the judge as a necessary step in reaching his or her conclusion in the sense that the judge would have reached a different conclusion if he or she had not thought it appropriate to express the ruling as broadly, or as narrowly, as it was in fact expressed. As Professor Neil Duxbury has observed in his insightful book The Nature and Authority of Precedent (2008), p 78:

It therefore seems to me that, when the ratio decidendi is described as a ruling or reason which is treated as "necessary" for the decision, this cannot mean logically or causally necessary. Rather, such statements must, I think, be understood more broadly as indicating that the ratio is (or is regarded by the judge as being) part of the best or preferred justification for the conclusion reached: it is necessary in the sense that the justification for that conclusion would be, if not altogether lacking, then at any rate weaker if a different rule were adopted."

Against that background we turn to consider whether the observations in Zeman on the FTT's jurisdiction are binding on us.

The structure of the decision in Zeman is relatively straightforward. The UT decision starts with five paragraphs of introduction, in the last of which the UT identifies the two issues which arise for decision: whether, on the assumption that the FTT had jurisdiction to deal with legitimate expectation, it erred in concluding that the appellant did not have a legitimate expectation on which it could rely. Secondly, the question whether the FTT has jurisdiction to consider public law issues in an appeal against a VAT assessment. The UT then summarised the background facts, the legislative background and the decision of the FTT. At paragraph [13] it turns to consider the first question, whether the appellant had a relevant legitimate expectation. At paragraph [19] it expressed its conclusion, that the appellant did not have a relevant legitimate expectation, and then at paragraph [20] it says,

In many respects, the structure of the decision in Zeman is very similar to that in Oxfam . In both cases there was a conclusion that the taxpayer had no legitimate expectation, and therefore its claim based on legitimate expectation was bound to fail, followed by an analysis of the jurisdiction issue.

As it simplest, the argument that the UT's conclusion on jurisdiction is a matter of decision is that the tribunal should only be using language which suggests that its conclusion on the legitimate expectation issue has led it to dismiss the appeal, if it has first concluded that it has jurisdiction to do that, a kind of chicken and egg point (if we dare use that expression in this case). Put another way, a positive conclusion on jurisdiction is a necessary first step in dismissing an appeal based on a lack of legitimate expectation, in contrast to declining to decide whether the tribunal has jurisdiction but adding that the claim based on legitimate expectation would fail even if it did. A purist view might be that the tribunal should not even do that; it should only hear argument on a matter and express any view at all if it is first satisfied that it has jurisdiction.

Unlike the FTT in Treasures of Brazil, w e do not consider that this is a case where the UT has given two alternative reasons for its decision, where both should be regarded as matters of decision. The positive conclusion on jurisdiction is not an alternative reason for deciding that the taxpayer's appeal in Zeman failed; a negative conclusion might be, but a positive conclusion is a necessary precursor to entertaining and deciding that ground in the first place.

As we have explained, at least on a purist view, the UT's conclusion on jurisdiction was a necessary first step to being in in a position to express a view on whether the appellant had a legitimate expectation, and so that conclusion was "necessary" to the UT's decision. But was that conclusion a "rule of law expressly or impliedly treated by the judge as a necessary step in reaching his conclusion, having regard to the line of reasoning adopted by him"? Although Nicola Davies and Haddon-Cave LJJ did not endorse Leggatt LJ's comments on precedent in Youngsam , they did not disagree with them. Moreover Nicola Davies LJ (with whom Haddon-Cave LJ agreed) considered that " ... the key words contained in the statement of Professor Cross are "treated by the judge as a necessary step in reaching his conclusion". " Leggatt LJ elaborates on that idea at [51], when he observes that "necessary" does not mean logically or causally necessary; it means what the judge considers "part of the best or preferred justification for the conclusion reached".

Looked at in that light, the only justification for the UT's decision in Zeman is that the appellant did not have a relevant legitimate expectation. At [19] they say that this conclusion is "sufficient to dispose of this appeal" and at [85] they again say that the appeal is dismissed "in light of our conclusion on the legitimate expectation issue". Indeed, in the same paragraph they say that their decision is reached in spite of ("notwithstanding") their conclusion on jurisdiction.

To the extent there are structural similarities between the decisions in Zeman and Oxfam , which might point towards the UT's observations on jurisdiction in Zeman being a matter of decision, we note that the UT in Noor , whilst it was inclined to move away from its views in Hok , deliberately did not express a concluded view on the point. In any event, the question for us is what was "treated by the [UT in Zeman ] as a necessary step in reaching [its] conclusion" and we are sure, for the reasons we have given, that, although the UT should arguably not have expressed a view on the legitimate expectation issue without first having reached a positive conclusion on jurisdiction, the only relevant justification the UT advanced for dismissing the appeal was that the appellant did not have a relevant legitimate expectation.

The conclusions of the UT in Zeman on the scope of the tribunal's jurisdiction are not binding on us.

Initially, we thought that we could deal with this question in quite short order. As the discussion at [341] above indicates, a decision is not reached per incuriam just because, if different arguments or different material had been placed before it, the court making the earlier decision might have reached a different conclusion. It applies only where the earlier court reached a decision in the absence of knowledge of a decision binding upon it or a statute, which, had the court had this material, it must have reached a contrary decision.

Mr Watkinson criticises the UT in Zeman for reaching its decision without canvasing all relevant authorities (in particular MIS ) and not considering the overall statutory scheme for tribunals in TCEA. Although Newey LJ delivered some fairly trenchant observations on the FTT's jurisdiction (or lack of it) to consider public law issues in MIS , that case was concerned with section 84(10) VATA, not the question of the tribunal's public law jurisdiction under section 83(1)(p). Although the way TCEA allocates judicial review jurisdiction to the UT alone is an important factor in considering the breadth of the FTT's jurisdiction, Zeman was concerned with the interpretation of a particular provision VATA, and the TCEA regime is not determinative of this issue.

It occurred to us as we wrote this decision that the provisions under which the appeals were brought in Corbitt and Zeman were near identical, and that prompted us to ask Mr Watkinson and Ms Sloane whether Corbitt is binding authority on the scope of the tribunal's jurisdiction under section 83(1)(p) VATA, so we are (and the UT in Zeman should have considered itself to have been) bound to hold that public law issues cannot be canvassed in an appeal under section 83(1)(p). If that is the case, Zeman would have been decided per incuriam.

Ms Sloane submits that, in Corbitt the House of Lords merely decided that the Tribunal did not have jurisdiction to supervise HMRC's exercise of discretion under Article 3(5) of the Value Added Tax (Works of Art, Antiques and Scientific Collections) Order 1972 (the "margin scheme"). Sales J in Oxfam addressed this issue. He considered that the House of Lords addressed the issue before it as a narrow point of construction, holding that since (as was common ground) the setting of requirements in a published notice within the first limb of Article 3(5) did not fall within the defined jurisdiction of the tribunal and there was no basis for distinguishing the position where the commissioners exercised their discretion under the second limb of Article 3(5), therefore as a matter of construction an appeal in relation to a decision under the second limb also did not fall within the jurisdiction of the Tribunal. Lord Lane's wider observation was obiter, as the House's decision was based on the narrow point just outlined.

In contrast, Mr Watkinson says that t he better view, on balance, is that Corbitt is binding authority that the tribunal has no supervisory/public law jurisdiction on a section 83(1)(p) appeal. He accepts that the precise focus in Corbitt was on whether section 40(1) FA 1972 permitted the tribunal to review the Commissioners' exercise of their discretion as regards the sufficiency of the records and accounts for the purposes of the margin scheme, but says that Lord Lane specifically considered both the assessment provision in section 31(1) FA 1972, and the appeal provision in section 40(1)(b), before observing that, if it had been intended to give a supervisory jurisdiction of that nature to the tribunal, he would have expected clear words to that effect in the Act. Instead, he observed, section 40 (1) set out nine specific headings under which an appeal may be brought "and seems by inference to negative the existence of any general supervisory jurisdiction".

He says that the justification for Lord Lane's conclusion would be weaker if his dictum as to the absence of the general supervisory jurisdiction was not adopted. Had he thought that section 40(1)(b) FA 1972 provided a supervisory jurisdiction, that would have justified a different conclusion. Lord Lane's final dictum was, therefore, part of the ratio of Corbitt, which includes that the Tribunal had no supervisory jurisdiction over an appeal against a VAT assessment.

He also points out that Corbitt has been treated as authoritative on this point in NatWest and Dollar Land and a similar approach (that the tribunal's jurisdiction is limited to the consequences of the statute and not the conduct of HMRC) was adopted in Marks and Spencer without express reference to Corbitt .

None of these cases referred to by Mr Watkinson concerned what is now section 83(1)(p) and it is therefore difficult to see them as confirming that Corbitt is binding authority on the scope of the tribunal's jurisdiction under section 83(1)(p). Indeed, despite his conclusion that Corbitt was "authoritative" that an appeal provision must be capable of being construed as giving the tribunal a supervisory jurisdiction, absent which the tribunal had no supervisory jurisdiction, in Dollar Land Judge J noted that the appeal provision (section 40(1) Value Added Tax Act 1983) had been expanded since the time of Corbitt and now included provisions which were accepted (including by HMRC) as supervisory in nature, concluding:

We are with Ms Sloane on this point. The very clear focus of the House of Lords in Corbitt was on the way the margin scheme worked and in particular the way the (accepted) lack of jurisdiction in the tribunal to entertain challenges to the accounting requirements imposed by the Customs and Excise Commissioners led to the conclusion that there could be no jurisdiction to entertain challenges to their decision not to recognise a trader's accounts (which did not comply with those requirements) as sufficient. Although Lord Lane commented that he would expect to see clear language in a statutory appeal provision before accepting that it conferred a supervisory jurisdiction, that observation was not necessary for his decision. He could have held that the tribunal had a supervisory jurisdiction on an appeal against an assessment and still reached the conclusion he did because of the way the margin scheme worked. Corbitt did not bind the UT in Zeman to come to a different conclusion.

The question of the breadth of the tribunal's jurisdiction is a vexed and difficult one, which has been considered in a number of previous decisions, but none of those authorities which Mr Watkinson identified as missing in action in Zeman (nor Corbitt , which was not on his list but which we raised ourselves and have just addressed) would have compelled the UT to a different conclusion, nor in fact did he specifically identify any which he thought would have that effect. They might have made the UT reconsider its view (although we rather doubt that given the range of authorities the UT did address), but it is not the case that any authority or statutory provision not addressed by the UT would have compelled it to reach a different conclusion.

The UT decision in Zeman on the jurisdiction question was not reached per incuriam.

We were not addressed to any great extent on the answer to this question. Mr Watkinson's position is that Zeman is wrong and, as it is not binding on us, we can and should say that it is wrong and decline to follow it. Ms Sloane says that, even if it is wrong, Zeman is binding on us. In argument, she said that, if we were to find (as we have) that Zeman is not binding on us, we should still follow it unless we were to conclude that it is "plainly wrong". Given the vexed and difficult nature of the jurisdiction question, her position is that, even if we thought Zeman was wrong on this point, we could not possibly think that it was "plainly wrong".

During argument, Ms Sloane drew our attention to a passage in Lewison LJ's judgment in Merck Serono , where he was addressing the question whether the CA should depart from a previous decision if it concluded that the earlier decision was decided per incuriam and was therefore not binding on it. He identified several points, arising out of the jurisprudence on when the highest court considers that it is appropriate to depart from an earlier decision of that court. Clearly, we are not the Supreme Court engaged in that exercise, but we respectfully found Lewison LJ's list of things to think about in that context to be helpful when deciding whether to depart from Zeman . His points (which start at [89]) are as follows:

Although not a case cited to us, the UT (a panel including Judge Greg Sinfield, then the President of this Chamber of the FTT) had this to say about horizontal stare decisis in the FTT in HMRC v Taher Suterwalla and another , [2024] UKUT 188 (TCC) at [23]:

This broadly reflects the summary of horizontal stare decisis in Halsbury's Laws of England (2020):

If the FTT should be slow to depart from a non-binding decision of an earlier FTT, the FTT should clearly be equally slow (if not slower) to depart from a non-binding decision of the UT reached after consideration on a matter arising out of a complicated and difficult enactment.

Our starting point, therefore, is that we should not depart from Zeman, even if we would have reached a different conclusion ourselves, unless we are convinced that it was wrong. Just as the UT in Suterwalla regarded "convinced" and "satisfied" as effectively the same test, we regard both of those as equating with whether we regard Zeman as "plainly" wrong. We can only be "convinced" that Zeman is wrong if we are so sure of that that we are prepared to say that it is "plainly" wrong.

Whether these are separate factors, or (at least to some extent) just aspects of our conclusion that Zeman was not "plainly wrong", our observations on how Lewison LJ's considerations play out here are as follows:

For the reasons set out above (particularly the need for the tribunal to speak consistently with one voice on an issue as fundamental as jurisdiction), we have decided that, although we do not consider that the comments on jurisdiction in Zeman are binding on us and although we are far from persuaded by them, we should follow them.

It follows that our answer to the second question (Does the FTT have jurisdiction to entertain Morrisons' legitimate expectation argument?) is "Yes".

Before we turn to consider the merits of Morrisons' legitimate expectation argument, we should briefly explain why we consider that the comments in Zeman on jurisdiction are wrong but not plainly so.

With the exception of Oxfam (which was not followed in Noor ), the authorities (including Corbitt, a House of Lords decision in relation to essentially the same appeal gateway as that considered in Zeman ) we have reviewed before Beadle lean heavily against the tribunal having a supervisory jurisdiction (to consider public law issues), even where the appeal is "in respect of" a decision which the tax authority "may" make. These conclusions have been reached in the light of the tribunal structure established by TCEA and the rule in O'Reilly v Mackman (that public law issues should generally be raised in judicial review proceedings in the High Court).

In Beadle itself the CA concluded that the tribunal had jurisdiction to consider an appeal against a penalty assessment, basing its decision on the accepted exception to the rule in O'Reilly v Mackman for "enforcement proceedings" except where (as was held to be the case there) the statutory scheme in question excludes the ability to raise a public law defence.

Nowhere in her judgment did Simler LJ equate an appeal against an ordinary tax assessment with enforcement proceedings, and assessment and enforcement are quite different matters. She was addressing a ground of appeal that, "in the absence of any statutory appeal process against PPNs, collateral challenges to PPNs on public law grounds may be made by a taxpayer in the course of a statutory appeal against a penalty notice for non-compliance with a PPN", put another way, the appellant was seeking to challenge the validity of the PPN in the course of challenging an enforcement action (the penalty) in relation to the unappealable decision to issue the PPN. That is radically removed from a statutory appeal against a tax assessment, and we are far from convinced that it is right (as the UT did in Zeman ) to equate a taxpayer bringing an appeal against a tax assessment with, in substance, a defendant in enforcement proceedings, and then go on to hold that the taxpayer should be entitled to challenge HMRC's decision to raise the assessment on public law grounds in proceedings before the FTT "unless that entitlement is excluded by the relevant statutory language".

Although, like Judge Brown KC in Drinks and Food, we entertain significant doubts about whether Zeman is correctly decided, there are clearly arguments (for example, those adopted by Sales J in Oxfam) to the contrary. Indeed, we note that the UT in Caerdav , although reaching the conclusion that a legitimate expectation argument could not be raised in that case and suggesting that the question of supervisory/public law jurisdiction should be approached from the opposite perspective (that the starting point is therefore that appeal grounds which concern public law arguments should be pursued in judicial review proceedings rather than before the FTT), appeared (at [155]) to endorse the decision in Zeman itself.

So, even if Zeman is wrong on jurisdiction, which we rather think it is, we agree with Ms Sloane that we could not possibly say that it is "plainly wrong".

Morrisons and HMRC have agreed that there are four issues to be addressed here, namely:

There is no real dispute as to the law in this area. There is a summary of the applicable principles in the context of VAT clearances in R (Airline Placement Limited) v HMRC , [2023] EWHC 1191 (Admin) (" APL "):

These comments largely reflect the classical statement of Bingham LJ in the Divisional Court in R v IRC ex p MFK Underwriting Agencies Ltd & Ors , [1990] 1 WLR 1545 (" MFK "), of what is required before guidance given to a taxpayer might create a legitimate expectation that HMRC would forego tax due under statute:

In APL (at [68]) Constable J described the test for materiality of non-disclosure as:

Appropriate (i.e. clear and unambiguous and devoid of relevant qualification) statements can be made by HMRC in dealings with individual taxpayers following a request for a ruling or in published guidance directed at the whole world, such as their published guidance booklet on individual residence ( R (Gaines-Cooper) v HMRC , [2011] UKSC 47 at [29]) or their Manuals ( R (Aozora GMAC Investment Ltd) v HMRC , [2019] EWCA Civ 1643 ).

Morrisons say that all these conditions are met, for the following reasons:

The first dealing between HMRC and Morrisons in relation to CDRCs was in the autumn of 2012. There is very little evidence about exactly what took place. We do know, from Officer Dean's letter of 5 September 2013, that he discussed the CDRC VAT liability issue with Mr. Prince and went on a site visit. His understanding of CDRCs is that they were "packaged in plain bags, left to cool naturally and made available (for self-selection by customers) from table type displays in the aisles". Following his store visit, Mr Dean said that he was satisfied that the tax treatment Morrisons adopted, that CDRCs were zero rated and HRC's standard rated, "appeared acceptable".

The prompt for Officer Dean's letter in September 2013 was his visit to a store, near to where his daughter lived, where he noticed that several aspects of the placement, promotion and sale of chickens had changed and that he found these changes concerning. He thought that it would be useful to follow up on his concerns to make sure that he had a full understanding of the position and "to minimise the risk of the group building up a substantial tax liability (and potentially penalties)".

It is clear from MFK that it is one thing to ask an HMRC officer whether he shares a taxpayer's view on a point and quite another to ask whether HMRC would forgo any claim to tax on any other basis. For us find that the discussions in 2012 are sufficient to establish a legitimate expectation, Morrisons must prove, on the balance of probabilities, that it was clear that HMRC were being asked for a formal ruling which they would stand behind. Similarly, Morrisons must show that they made full disclosure to HMRC of all relevant material and the statement Officer Dean gave must be clear, unambiguous and devoid of relevant qualification. If these conditions are met, an ordinary sophisticated taxpayer could reasonably conclude that (whatever HMRC actually thought) HMRC had created a legitimate expectation.

All we know about what took place in the autumn of 2012 is that there was a discussion and a site visit which led to Officer Dean concluding that what Morrisons were doing "appears acceptable", which is consistent with his comment in his witness statement that "there were no issues of contention". We do not know what Morrisons asked him for or told or showed him or indeed in exactly what terms he delivered his thoughts to Morrisons.

Whilst (of course) a ruling can be given orally, in the absence of evidence beyond what we can learn from Officer Dean's letter and witness evidence (which absence itself rather suggests to us that Morrisons were not seeking a ruling on the VAT liability of CDRCs - if they were, we consider that they would have generated records of their dealings with HMRC and kept them safe), we do not consider that there is evidence of dealings from which an ordinary sophisticated taxpayer could reasonably conclude that HMRC had created a legitimate expectation.

During his submissions, Mr. Watkinson described the dealings between Morrisons and HMRC as "fireside chats". Ms Sloane says that that rather underestimates their seriousness and points to the acceptance in this letter by Officer Dean that getting the VAT treatment of CDRCs right is important for Morrisons in order to avoid them building up a significant VAT liability. We agree with Ms Sloane that the discussions in 2012 (even though we do not know much about them) and the correspondence in 2013/14 clearly amounted to a serious dialogue, but that is not the same as saying that they concluded with something from which an ordinary sophisticated taxpayer could reasonably conclude that HMRC had given an unequivocal ruling on VAT liability. Indeed, there is absolutely no evidence of Morrisons asking HMRC for a "ruling" (a confirmation that they regard CDRCs as zero rated and will forego any claim to tax on any other basis) as opposed to asking Officer Dean whether he shared Morrisons' view of a legislative provision. As MFK makes clear, those two things are miles apart.

Moving on to the correspondence in the autumn of 2013 and the early part of 2014, this starts with Officer Dean's letter of 5 September 2013. This starts with a summary of what took place in the autumn of 2012, including what he saw on his site visit in 2012.

After that, the letter records Officer Dean's concerns from visiting (in a personal capacity) the store near where his daughter lived. He acknowledged that this was an important issue both for HMRC (so he wants to understand what Morrisons are doing) and for Morrisons and suggests a meeting to take the matters forward.

In her email of 17 September 213, Ms Harrison asked Officer Dean to "be more specific as to why you feel the VAT liability of the cool down chickens has been changed". Mr. Watkinson says that this email amounts to an acknowledgement that, if Officer Dean had made a ruling in the autumn of 2012, this had now been withdrawn. Given our conclusion that there was no ruling sufficient to establish a legitimate exhibition expectation in 2012, this is an academic point. We agree, however, that Ms Harrison realised that Officer Dean's view had changed, although she did not know why.

Officer Dean replied to Ms Harrison on 20 September 2013, as set out at [52] above. It is clear from this email that Officer Dean was clearly not giving an unequivocal ruling in relation to the VAT liability of CDRCs. In the second main paragraph of his email he indicates that Note (3B)(a) may well need to be explored in some detail in the future. Clearly, that reservation alone indicates that he was not signing off on all five tests in Note (3B), and there is nothing from which an ordinary sophisticated taxpayer could reasonably conclude that Officer Dean was giving an unequivocal ruling on the overall VAT liability of CDRCs. Apart from any specific points he made, this was just the beginning of the correspondence, so any "sign off" would have been wildly premature.

Officer Dean then refers to his summary of the initial position regarding CDRC displays in his previous letter, but now he runs through what he saw in 2012 and what he observed more recently. He records his initial understanding that CDRCs were placed together for display and self-selection by customers from a table several metres away from the area housing the ovens and rotisserie counters. The tables were of a plain type and not heated and the bags were "essentially plain paper (with barcode and VAT free stickers)". So, he commented, the chickens were not displayed or sold in heat retentive packaging within Note (3B)(d). There was no advertising/signage on the packaging, the display tables or the immediate environment to indicate that the products were supplied hot. He concluded that the group had "clearly structured the arrangements to differentiate the "VAT free/cool down" chickens from HRCs and "could support a case" that Note (3B)(a) was not engaged.

He then says that he observed on his recent visit that CDRCs were no longer being displayed at some distance from the rotisserie on tables. They were displayed for self-selection from a stainless-steel open display next to the heated rotisserie display counter. This might engage Note (3B)(c). He said that there were "no changes regarding the immediate packaging"– so the packaging used is not breaching Note (3B)(d).

His main concern related to advertising and signage in potential breach of Note (3B)(e). This, of course, formed a large part of the subsequent discussions.

Overall, he was concerned, given "a number of indications ... that "cool down" products are being re-integrated into the main rotisserie range, [that] the entire approach of the group to these sales, in terms of determining it's (sic) "purpose" - under the Note 3B a) test - may need to be examined".

Officer Dean wrote a letter to Rowan Robinson (VAT manager - Taxation at Morrisons) on 15 January 2014. This followed his conversation with colleagues in the RUE. He referred to the objective tests introduced by FA 2012 and the group's concern about how phrases used within the example signs they had provided could impact on VAT liability. Having introduced the signage issue, Officer Dean says, "I would stress that my response to your request purely addresses is that single issue" (i.e. whether the proposed signage/placement means that products are being advertised or marketed as hot within Note (3B)(e)). He goes on to say that:

The rest of his letter is devoted to a detailed analysis of signs he has been sent. He concludes that, "based on assurances given by the group with regard to intention and in the absence of other indicators" of standard rated treatment and despite his still harbouring concerns, applying a blanket standard rate treatment to products sold under the signs he has seen and approved "would not be (fully) justified". He signs off by observing "Clearly, any decision as to the precise format/wording of any marketing/advertising material adopted is a matter for the group but I do hope these comments are useful in informing your deliberations."

On 11 September 2015 Morrison's wrote to HMRC making a voluntary disclosure of underpaid VAT because of wrongly categorising certain products as zero rated. In the introductory part of that letter they referred to the 2012 changes affecting the liability of hot takeaway food and said that, as a result of these changes, all hot products sold from the "Oven Fresh" counter became liable to VAT at 20% with the exception of "VAT free cooldown chickens which meet the new conditions for rating". We have seen Mr. Nichols point to this letter, but this letter is primarily explaining the (very significant) VAT coding errors Morrisons has found and admits to. The VAT treatment of CDRCs is presented as part of the background picture and no more. It would be wholly unreasonable to read this letter and HMRC's not replying and questioning the VAT treatment of CDRCs as any kind of acceptance by HMRC that CDRCs were being correctly zero-rated. HMRC are not asked to say anything about CDRCs, and therefore not surprisingly they pass no comment on them.

Mr. Watkinson criticises Morrisons for (he alleges) trying to assemble a VAT ruling out of several component parts, in other words taking passages or exchanges from different pieces of correspondence or other dealings and adding them together to make a ruling. He says that there is no authority that this can be done, and the clear implication is that, if HMRC are being asked to rule on the VAT liability of a product, they must be asked, in clear terms, to do exactly that taking all factors into account at the same time. It is, he says, not possible to assemble a ruling piecemeal or by stealth. Put in those terms, that must be right; MFK makes it abundantly clear that any representation which is sufficient to establish a legitimate expectation must be obtained from HMRC in response to the taxpayer making it abundantly clear why they are asking a particular question and what they are going to do with the answer. That said, in a case (like this one) where several, quite distinct, criteria need to be considered, we can see no fundamental reason why a taxpayer in Morrisons' position might not ask HMRC for a clear ruling on one issue and then come back later and ask for a clear ruling on another, as long as they make it clear that they are asking for a ruling in each case and indicate what the effect of a later ruling would be, when combined with rulings already obtained.

The relevance of that point here is that, as we have just seen, the correspondence in the autumn of 2013 and the early part of 2014 is focused primarily on Note (3B)(e). To the extent Officer Dean made any comments on Note (3B)(a), (c) or (d), they were made in 2012 and merely repeated in his letters of 5 and 20 September 2013.

To be completely clear on this point, for the reasons discussed at [401]-[404] above, we do not regard Officer's Dean's comment that what he saw on 2012 "appeared acceptable" to be a clear, unambiguous statement devoid of relevant qualification, which an ordinary sophisticated taxpayer could reasonably take to be a ruling by HMRC on the VAT liability of CDRCs, either in the Autumn of 2012 or when it was repeated on 5 September 2013.

If we look for a moment at what Officer Dean said in his letter of January 15 2014, the first point to make is that he was very clear that he was addressing Note (3B)(e) only. He was not even providing a complete commentary on Note (3B)(e), as the examples he gives of other issues that would need to be thought about, such as whether pies and pasties put in cool down areas might be the subject of other in store advertising indicating a supply whilst hot, make clear. All he was doing was passing comment on particular signs, and whether products sold adjacent to or underneath them would eo ipso be treated as advertised or marketed as hot for the purposes of Note (3B)(e).

Even though Officer Dean makes some comments in relation to those signs, he is very clear at the end of his letter that the decision on the signs to use is for Morrisons to make and that he hopes his comments are "useful". He was very definitely not saying anything more expansive; indeed, given his concerns about the potential effect of "other circumstances", it is hard to see how he could do that. At most, he was saying, by way of (he hoped) helpful comment, that, if the only issue to be considered in relation to Note (3B)(e) was these signs (i.e. absolutely nothing else pointed to standard rating), he would not expect ("in the vast majority of cases, I do accept") that the signs would cause the product in question to be standard rated. There is nothing here from which an ordinary sophisticated taxpayer could reasonably conclude that Officer Dean had created a legitimate expectation as to the VAT liability of CDRCs generally or in relation to Note (3B)(e).

Going back in time to the longer email of 20 September 2013, again there was no complete sign off on the liability of CDRCs.

So, if we look at what Morrisons had obtained from HMRC by the start of the assessment period (in January 2017) in relation to the tests in Note (3B), we can see:

We cannot read into the very small amount of coverage of Note (3B)(c) or (d) in September 2013 anything that could reasonably be taken to be a ruling by Officer Dean on the position of CDRCs so far as these two tests are concerned. We can see Officer Dean's concern about Note (3B)(c) and heat transfer from adjacent cabinets. We know that this concern was allayed, but there is no record of the outcome in this correspondence. As far as packaging is concerned, the correspondence records what Officer Dean saw and his assumption, as he thought he was looking at essentially plain paper bags which were not heat retentive, that the packaging raised no issues as regards Note (3B)(d). The focus of the 2013/2014 correspondence was on advertising and marketing, ultimately only on the impact of the wording of Morrisons' signage on Note (3B)(e). That was where Officer Dean's attention was ultimately focused. His "initial/current" summary (in his email of 20 September 2013) was (as he put it) to set the issues he was concerned about (signage/place of display) in context, given Ms Robinson's request for "early advice" so Morrisons could think about the triggers that had raised his concerns. In the light of those comments and the explicit limitations in Officer Dean's letter of 15 January 2014, we do not consider that a reasonable sophisticated taxpayer reading this correspondence would distil a ruling on chicken paper bags and Note (3B)(c) and (d) from the references to this issue in correspondence where the writer's attention was focused elsewhere and where he ultimately ended up giving only guarded guidance on the issues that were "front and centre" in the correspondence..

We do not know why Officer Dean thought that the chicken paper bags were plain paper, and therefore non-heat retentive, bags. We know, from Mr. Whitaker's evidence if nothing else, that all packaging is heat retentive to some degree and that the chicken paper bags perform well so far as heat retention is concerned. We also know that, as late as 2021, Morrisons were telling HMRC that the chicken paper bags were not heat retentive, when, as Mr Nichols readily acknowledged in cross examination, this was just not true.

We do not know whether Morrisons told Officer Dean anything about the bags in 2012, which he repeated in his 2013 emails. Ms Sloane says that he could see them for himself, but it is not Officer Dean's job to try to work out the properties of the chicken paper bags even if (which we very much doubt) a non-expert could work out how heat retentive these bags were just by looking at them. Mr Whitaker did find out how (absolutely and relatively) heat retentive the bags are, but he and his team conducted a significant amount of research before reaching their conclusions. If Morrisons wanted a ruling on the VAT liability of CDRCs, generally or on particular aspects of the new tests introduced in 2012, it was for them to put their cards face up on the table. There is no evidence of Officer Dean being told anything about the bags at all; we know no more than that he saw some CDRCs sitting in these bags on a table.

We do not need to know why Mr Prince and his colleagues did not (so far as we know) tell Officer Dean anything about the heat retentive properties of the bags in 2012, or did not correct his misapprehension when he repeated it in his 2013 emails, or made an incorrect statement to HMRC in 2021. (2021 is, of course, after the end of the assessment period and statements made by Morrisons in 2021 are not relevant to the question whether Morrisons had a legitimate expectation at the start of the assessment period.) Mr Nichols said that Morrisons would never try to deceive HMRC, and (we should stress) no one has suggested, even for a moment, that they did. There is evidence of mis/non-communication between the Tax Department at Morrisons and other parts of the group. Mr Watkinson described the relationship between the Tax Department of Morrisons and the operational parts of the group as "chaotic". There is insufficient evidence to suggest that the relationship was systemically chaotic, but it is undoubtedly the case that internal communication was imperfect from time to time. That may well explain that failure. Another explanation (the possibility of which was touched on during Mr Nichols' cross-examination) is that no serious research had been carried out and Morrisons simply did not know about the heat retentive properties of the bags, and they proceeded (and allowed HMRC to proceed) on an important, but wrong, assumption they had not rigorously checked.

What is abundantly clear, though, is that, if HMRC had known about the heat retentive properties of the chicken paper bags in 2012/13, there is more than a "real possibility" that Officer Dean would not have made the comments he did. We hold that there was material non-disclosure so far as that issue was concerned. If (which we do not consider to be the case at all) Officer Dean's comments so far as Note (3B)(c) or (d) and the chicken paper bags were concerned would otherwise amount to a representation which could establish a legitimate expectation, Morrisons' material non-disclosure (whatever its explanation) of the heat retentive properties of the chicken paper bags would negate that conclusion.

Officer Dean also said that he was concerned that he had not been made aware of the two-hour limit on the time Morrisons could expose CDRCs for sale. He told us that this would have raised concerns as, contrary to his understanding (which he says he obtained from his discussions with Mr Prince in 2012), CDRCs could not fully cool down to the ambient temperature before sale. We have already expressed (at [230]) our conclusion that a policy of taking CDRCs off sale after two hours at a time when they are still well above the ambient temperature, is not a problem so far as Note (3B)(c) is concerned, as that does not (in our view) amount to a step taken to ensure that particular CDRCs remain hot within Note (3C). But we have also seen that the process of construing these provisions is not straightforward and that our reading does not reflect Mr Gauke's policy explanation. For those reasons, we consider that this factor (that CDRCs could only be sold hot and would never fully cool down to or below the ambient temperature whilst on sale) was a "legitimately relevant" ( APL at [74]) issue for Officer Dean to consider. He should have been given an opportunity to reflect on that issue, given that what Morrisons were doing ran counter to Mr Gauke's articulation of the policy behind the legislation and the concerns Officer Dean had articulated about "the entire approach of the group to these sales" and the impact on Note (3B)(a). Given that, there was a "real possibility" that, had he known of this, he might have refused to give Morrisons any kind of guidance (which we consider to be just as relevant as the possibility that he might give a ruling or commentary in different terms) about CDRCs at all or formed a (or a different) view on Note (3B) (a) or (c) which negated any reassurance (if such it was) he gave on Note (3B)(e). If (which we do not consider to be the case at all) Officer Dean's comments would otherwise amount to a representation which could found a relevant legitimate expectation, Morrisons' material non-disclosure of the requirement to take CDRCs off sale after two hours (while they were still well above the ambient temperature) would negate that conclusion.

For the reasons we have just given, we do not consider that there is anything in HMRC's dealings with Morrisons which is sufficient to establish a legitimate expectation either that the overall VAT treatment of CDRCs was that they would be zero-rated or that HMRC would take a particular position as regards any aspect of the storage or marketing of CDRCs and any of the tests in Note (3B).

We turn now to look at HMRC's published guidance, to see if there is anything here which could found a legitimate expectation as to the VAT treatment of CDRCs. This guidance comprises VAT Note 709/1 (from which we have set out some extracts earlier in our decision) and relevant passages in HMRC's Manuals.

So far as HMRC comments on the particular tests in the passages in paragraph 4.3 of VAT Notice 709/1, are concerned, we do not find anything here to be particularly helpful to Morrisons. The commentary on Test 3 (whether food or drink has been kept hot) focuses on whether products have been stored in cooling down ovens or other appliances that slow down the rate of cooling, and all the examples involve appliances (such as heat lamps or hot water baths) of this nature. There is, however, no comment to the effect that it is only the use of an appliance which engages this test.

The commentary on Test 4 (the packaging in which products are supplied) restates the legislation and then says that in practice this will mainly affect products that are supplied in specialised packaging, such as foil lined bags and insulated containers, and that it will not affect products that are sold in ordinary paper bags or similar packaging. To the extent that specialised packaging might include anything beyond an ordinary paper bag, this comment is unhelpful to Morrisons, as is the specific confirmation about ordinary paper bags or similar packaging. As we have seen, the chicken paper bags are not foil lined, but they are not ordinary paper bags either. One of the examples under Test 4 is, however, very unhelpful to Morrisons. It refers to cooked chickens being supplied in heat retentive packaging or packaging designed to prevent the leakage of hot fluids or grease. We know that the chicken paper bags, although not designed to retain heat, are heat retentive and, much more importantly, are designed to prevent the leakage of hot fluids and grease.

So far as Test 5 is concerned, there is clear confirmation that phrases such as "freshly baked" do not indicate the products are supplied hot, but there is confirmation, if it were needed, that advertising or marketing a product as hot could be by way of a picture (of a product with steam coming off it) not just by words saying that it is hot.

Paragraph 4.4 of Notice 701/9 contains examples of hot food products that are caught (or not) by the new tests. We see the text starting with a warning that the new tests ensure that the vast majority of hot takeaway food is standard rated. The exception is hot food that does not satisfy any of the tests in Note (3B) and it then (perhaps echoing Mr Gauke's comment) gives as an example freshly baked bread that is "incidentally hot" at the time it is sold. CDRCs, as we have seen, are very much not "incidentally hot" at the point of sale.

If we turn to look at Example 5 in paragraph 4.4, we again see a discussion of a hot freshly cooked chicken being standard rated just because it is provided to customers in "specially designed (foil lined) bags designed to prevent leakage of fluids and grease" even though the chicken is fine as regards the other tests. There is a clear warning here, that selling hot chickens in packaging designed to retain fluids and grease is sufficient on its own for the chickens to be standard rated. The text here refers to "(foil lined)", but there is nothing here that justifies a business taking the view that only foil lined bags are a problem and that any other "bags designed to prevent leakage of fluids and grease" from hot food do not create problems so far as zero rating is concerned.

Officer Dean, of course, passed absolutely no comment on whether the chicken paper bags retained grease and fluids. He only said that he (incorrectly) thought they were ordinary paper bags that did not retain heat. Morrisons, of course, knew that the chicken paper bags were specifically designed to retain grease and fluids. The clear evidence from Mr Maestri is that the bags were designed to be, and from Mrs Whittle that they needed to be, leakproof. This is something else Morrisons did not share with Officer Dean.

The extract we looked at from the VAT Food Manual (VFOOD4260) deals with the purpose text (in Note (3B)(a)) and tells us that HMRC consider it to be "necessary to consider all the facts and circumstances relating to how the food is supplied" if it is suggested that food supplied hot does not meet this test.

Except for the confirmation that marketing something as "fresh" is not the same as marketing it as "hot" (which is relevant for Note (3B)(e)), we cannot see anything in VAT Notice 701/9 or this passage from the Manual that helps Morrisons to establish a relevant legitimate expectation. The commentary on Note (3B)(a) reflects, but does not add to, what can be derived from the caselaw. The comments on Note (3B)(c) are not unhelpful to Morrisons, in that they focus on keeping products hot using appliances, but the text is clear that this test will "in practice ... mainly affect" products kept hot this way; there is no suggestion that its effect is restricted to products kept hot using appliances. Most importantly, the commentary, so far as it touches on Note (3B)(d), is very unhelpful. An ordinary sophisticated taxpayer who knew what VAT Notice 701/9 had to say about packaging could not reasonably construe what little Officer Dean had to say about chicken paper bags as a ruling that they were not problematic as far as Note (3B)(d) was concerned.

For the reasons we have just given, we do not consider that there is anything in HMRC's published guidance which is sufficient to found a legitimate expectation either that the overall VAT treatment of CDRCs is that they would be zero-rated or (beyond the helpful point that marketing a product as fresh is not to be equated with marketing it as hot and the unhelpful point about grease retentive packaging) that HMRC would take a particular position as regards any aspect of the storage or marketing of CDRCs and any of the tests in Note (3B).

Our answers to the first two of the four questions set out at [394] above are:

Our answers to these questions make it unnecessary for us to prolong an already extremely long decision by considering questions (3) and (4).

For the reasons discussed above we have concluded that:

For the reasons set out above, we have found against Morrisons on both the Liability Ground and the Legitimate Expectation Ground.

This appeal is dismissed.

This document contains full findings of fact and reasons for the decision. Any party dissatisfied with this decision has a right to apply for permission to appeal against it pursuant to Rule 39 of the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009. The application must be received by this Tribunal not later than 56 days after this decision is sent to that party. The parties are referred to "Guidance to accompany a Decision from the First-tier Tribunal (Tax Chamber)" which accompanies and forms part of this decision notice.