J F Hillebrand UK Ltd v Revenue & Customs

Mr Saward explained that Hillebrand transports a very large number of consignments of goods for Tesco, so many that it has access to Tesco’s computer records of receipts in order that Hillebrand may make from them the entries to EMCS which are required. Ordinarily, said Mr Saward, the Hillebrand clerk making the entries accesses Tesco’s computer within about 24 hours of the arrival of the goods. It is clear that Hillebrand first learnt that Tesco had recorded receipt of only 1800 cases when its computer record was interrogated. There was no record in this case of when precisely the clerk undertook that task; Mr Saward suggested that some of the delay may have occurred because the clerk was on holiday, but I do not think the reason matters much for present purposes.

On 27 April a member of Hillebrand’s UK staff sent an email to its Spanish associated company stating that Tesco had claimed that the load was 400 cases short, and asking that a check be made with Freixenet. The response, on the same day, was “Freixenet has confirmed 2200 boxes were loaded”.

The next development of which there was evidence available to me occurred on 3 May 2011 when Hillebrand wrote to HMRC at Glasgow, identifying the consignment, and stating that

It was put to Mr Saward that the letter (which he did not write) was not entirely candid in that it did not mention the email exchange of 27 April, and the supplier’s response, and Mr Saward accepted that the letter might have been better expressed.

On 5 May 2011 the requisite entry was made to EMCS, to record the discharge. I deduce that the lateness was due, at least in part, to the fact that there was some uncertainty about how many cases had arrived. I had no explanation of the entry on the EMCS record of the date of arrival of the consignment as 3 May, but it seems to me that it was probably due to simple error. One of the fields to be completed on the screen before the completed entry is submitted requires the entry of a code, of which there are several, though only two are relevant here: 1, which signifies “Receipt accepted and satisfactory” and 2, “Receipt accepted although unsatisfactory”. It is undisputed that the entry actually recorded on EMCS used the code 1. I should add, if only for completeness, that it is not possible to amend an entry once it has been submitted. It seems that the discrepancy between that entry and the letter of 3 May led to HMRC’s realisation that there was, or potentially was, an under-payment of duty since, as Hillebrand said in the letter (even if somewhat confusingly) it had not paid the duty on the 400 cases.

Mr Saward told me that Hillebrand uses its own bespoke software which forms an interface between Hillebrand’s systems and EMCS, and he produced some screenshots showing the entry in the course of being made, and before submission. These screenshots show that the code 2 has been selected, and that a further sub-screen has been opened to record that the reason why the receipt was unsatisfactory was that there was a shortage. However, it is not possible to determine when and for what purpose the screenshots were taken, and it is quite possible they are later re-creations, rather than contemporaneous. Although code 1 is entered on EMCS, Hillebrand’s own record, generated by the same software, shows that code 2 has been entered. Mr Saward asked the software suppliers whether the various entries made could be traced in order to provide an audit trail as a means of determining how the mis-match had occurred but was told, in an email produced to me, that it was not possible.

Mrs Hardy’s evidence was that, had the software house been able to offer an explanation of the error she might well have accepted it but, in the absence of any explanation, she felt obliged to regard the EMCS entry as accurate. Nevertheless, she decided to make some further enquiries before raising the assessment which is the subject of this appeal.

On an unknown date (since none is recorded on it) but which was plainly in late 2011 Mrs Hardy arranged for a mutual assistance request to be sent to the Spanish tax authorities, asking that the size of the consignment be checked. The response, also undated, but sent to Mrs Hardy by a colleague in Glasgow on 9 January 2012, reads as follows:

The outcome of that enquiry led Mrs Hardy to the conclusion that she could not be satisfied that the load had consisted of only 1800 cases, and she raised the disputed assessment.

Soon after, Mr Saward made some enquiries of his own of Tesco. He received an email dated 9 May 2012 from a person described as an accounts executive in Tesco’s indirect tax and duty office, stating that “The supplier was paid for 1800 cases only”. Included in the email was a screenshot of Tesco’s accounting record, showing an entry dated 23 May 2011 with 2200 entered in a column headed “quantity” and with a figure in a column headed “cost”, a further entry of the same date of -400 in the “quantity” column and a negative figure in the “cost” column, and a third entry of 3 August 2011 showing zero in the “quantity” column and a figure matching that in the previous line, but positive instead of negative, in the “cost” column.

Mr Saward told me that the Milton Keynes depot had now closed, and that he had since made further enquiries of Tesco, the result of which was available to me. Tesco produced a copy of their stock listings, showing purchases from Freixenet, including this consignment; it show an order for 2200 cases and the receipt of 1800. The covering email states that there were no stock adjustments, save for three cases which were added and then almost immediately removed again, and no occasion on which Freixenet were shown as having delivered more cases than were ordered (which might account for the missing cases). Mr Saward added that no discrepancies had been found when the warehouse was closed and the remaining stock removed. Mrs Hardy, however, told me of other instances in which Tesco had made mistakes; although the errors had been resolved they nevertheless were errors, and she could not rule out the possibility of such an error on this occasion when the supplier was adamant that the full quantity had been sent. In addition, she was aware that goods off loaded by Tesco from a container were laid out on the warehouse floor before counting, which did not always occur immediately, and it was possible the 400 cases had been removed to the storage racks before they were counted in this case. Mr Saward retorted that such an eventuality was unlikely as goods were always put in the racks before being removed for dispatch to stores or other warehouses.

It is common ground that the burden of showing that the duty is not payable—which in the circumstances of this appeal is the equivalent of showing that the 400 cases did not arrive—rests on Hillebrand, by virtue of s 16 of the Finance Act 1994. The standard is the balance of probabilities, and the question I must resolve is whether that burden has been discharged by the evidence I saw and heard, unsatisfactory (since it is incomplete) as it is.

It is, of course, human to err and I start from the proposition that either Freixenet or Tesco could be mistaken. In favour of the proposition that it is Tesco which has erred are the facts, first, that there is no evidence before me of an error by Freixenet in any other supply to Tesco transported by Hillebrand; that it has been asked by Hillebrand itself to confirm the quantity of stock in the consignment and has asserted that 2200 cases were sent; and it has been asked a similar question by the Spanish tax authorities. Nevertheless, I am bound to say I have some concerns about the quality of that evidence.

I was surprised to see that the document produced, as the response from the Spanish authorities and not simply a translation of it, was on plain paper and in flawless English. The mutual assistance responses I have seen in the past have been on official stationery, and written either in the writer’s own language or in less than perfect English. It is, of course, entirely possible that the writer of this response does have an excellent command of English and I do not read much into that factor. But it is also surprising, and in my experience unusual, to see that the writer’s name has been redacted. Additionally, there is nothing to indicate the standing within Freixenet of the source of the information, nor how thorough the checks were; “no differences have been detected” may mean no more than that nothing has come to light, and does not imply a positive check. The statement that the shipper is “not aware of any incident” would be untrue if Tesco had paid for only 1800 cases rather than the 2200 supplied, since the reduced payment can scarcely have gone unnoticed.

On the other hand, it is not immediately apparent why Freixenet should claim that the full consignment had been shipped if it had in fact been paid for only 1800 cases. It is conceivable that there is, or was, a dispute between Freixenet and Tesco and Freixenet felt it had to maintain its position that the full consignment had been sent; but this possibility, which is in any case speculative, does not explain why Freixenet should state that it was unaware of any incident.

The evidence obtained from Tesco, too, is not entirely satisfactory; it would, in particular, have been helpful to see the remittance advice, or more probably its electronic equivalent, by which Tesco informed Freixenet that it was paying for only 1800 cases, and any response. Mr Pritchard suggested that the screenshot produced by Tesco was unreliable because the third line of it as I have described it above might indicate that the missing 400 cases had been found, and that Tesco had after all paid for them. I do not accept that argument; the screenshot was quite clearly included in the email by its writer as the evidence supporting her statement that payment had been made for only 1800 cases. Mr Pritchard and I may struggle to understand a Tesco accounting document, but I find it impossible to believe that a member of Tesco’s accounts staff would make a statement supported by a document which revealed to anyone who could understand it that her statement was untrue. I am satisfied by this evidence that, by a date more than a year after the consignment was dispatched, Tesco had paid for only 1800 cases.

The evidence showed that errors had occurred in Tesco’s warehouse in the past. Mrs Hardy described a cluster of errors which had take place shortly before the time at which this consignment had arrived, and which had resulted in management intervention. She said that the errors had apparently been resolved at that time, but thought this might have been a recurrence, albeit isolated. She was critical of Tesco’s method of removing goods from a container but not immediately checking the quantity and, in correspondence between the parties during the course of the dispute, it was suggested that Tesco should have carried out a stock check, in order to verify that the missing cases were not, after all, in the warehouse. In my view that was an unrealistic suggestion; as Mrs Hardy and Mr Saward agreed, the warehouse was vast and the undertaking of a full stock take, in order to discover whether or not what (by the standards of the warehouse) was a small quantity of goods was present, would have been disproportionate. I mention again that there is no evidence that a similar request was made of Freixenet.

I accept, as did Mr Saward, that code 1 (“Receipt accepted and satisfactory”) was entered to EMCS, but I am satisfied that was a mistake. I place no reliance on the screenshots which Mr Saward produced, since it is not clear they are contemporaneous. But I find it difficult to accept that Hillebrand would write to HMRC on 3 May to report a short delivery, only to make, intentionally, an entry reporting a full delivery two days later, and I reject that as a possibility. I attach some weight to the fact that Hillebrand’s own records show that code 2 was entered, but as I had no evidence of the manner in which its software functioned I do not regard this point as decisive. However, it should be remembered that what Hillebrand entered to EMCS is merely evidence of what its staff understood at the time, which was no more than what Tesco’s records showed; it is not evidence which helps me decide whether the counting error lay with Freixenet or Tesco.

As I have said, it is for Hillebrand to satisfy me, on the balance of probabilities, that the 400 cases were not loaded into the container in Spain and taken out of it in Milton Keynes. It will be apparent from what has gone before that the evidence available is of poor quality, and that much of it can be criticised. For those reasons I have not found this an easy case. However, I am persuaded that the weight of the evidence favours Hillebrand, and that it has discharged the burden.

In reaching that conclusion I accept that the Tesco stock records are consistent with there being a shortage, and with there being no late discovery of an otherwise unexplained excess of stock. I accept too the evidence that Tesco paid for only 1800 cases; for the reasons I have given I reject the argument that the third line of the relevant screenshot might show a late payment. Both of those pieces of evidence support the proposition that the delivery was short. The evidence from the supplier, by contrast, amounted to nothing more than assertion; it was backed up by no stock records, nor any evidence that Tesco had paid in full. In addition, as I have said, the contention that the supplier was unaware of any incident is difficult to understand; even if, on Mr Pritchard’s hypothesis and contrary to my finding, Tesco paid initially for 1800 cases and later for the additional 400, the supplier ought to have been aware that something untoward had happened. I bear in mind that there is some evidence of mistakes by Tesco (albeit mistakes which were corrected), and none of other mistakes by Freixenet, but in my judgment, on the evidence available to me, it is nonetheless more likely that the error lay with Freixenet than with Tesco.

For those reasons I have concluded that the appeal is to be allowed, and the assessment discharged.

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.