VALUE ADDED TAX Restaurant Observations and Test Purchases Whether Reliable Evidence Yes

VALUE ADDED TAX — restaurant — observations and test purchases — whether reliable evidence — yes — inferences drawn from witnessed cashing-up — takings much higher than on any other day — proportion of credit card sales to total materially different from any other occasion — whether reliable test — yes — no dishonesty penalty imposed — whether indication that evidence of suppression lacking — no — assessment made but not notified then reviewed and entirely recalculated by another officer — whether first assessment survives — no — later assessment partly out of time — appeal dismissed in principle but adjustments to assessments to be made

MANCHESTER TRIBUNAL CENTRE

GOLDKEY RESTAURANT LIMITED

Appellant

- and -

THE COMMISSIONERS OF CUSTOMS AND EXCISE

Respondents

Tribunal: Colin Bishopp (Chairman)

Marjorie Kostick FCA ATII

Sitting in public in Manchester on 15, 16 and 17 September 2003

Leslie Allen of counsel, instructed by Ernst & Young, for the appellant

Clare Grundy of counsel, instructed by their solicitor’s office, for the respondents

© CROWN COPYRIGHT 2003

1

DECISION

1This is an appeal by Goldkey Restaurant Limited against two VAT assessments. The first, for £66,319 and interest, was notified on 31 January 2000 and it covers the consecutive accounting periods from 1 July 1996 to 31 December 1998. The second, for £15,115 and interest, was notified on 21 February 2000; it covers the two accounting periods running from 1 January to 30 June 1999.

2A notice of appeal was served on the appellant’s behalf by his then representatives who, according to the date endorsed on the notice of appeal, despatched it to the tribunal on 10 February 2000. It may be that the appellant or its advisers realised that a second assessment was likely, but it obviously could not have been included in that notice of appeal. It is referred to in the respondents’ statement of case, served in September 2000. Although the wording of the statement of case is a little ambiguous, one forms the impression that its draftsman had approached his task upon the assumption that both of the assessments were challenged. However, the notice of appeal has not been amended, nor has a second notice of appeal been submitted and, therefore, no appeal against the second assessment has been mounted. This omission was pointed out to us by Leslie Allen, counsel for the appellant, at the outset of the hearing. He was unable to offer any explanation of the oversight, but asked us to direct that the appeal should be amended, albeit well out of time, so as to include the second of the assessments.

3His application was opposed by Clare Grundy, counsel for the respondents. She argued that, were we to allow the application, the appeal against the second of the assessments would be more than three years out of time, that the matter had been before the tribunal on two previous occasions when the error had not been noticed, and that the Commissioners should not be prejudiced by an error on the part of the appellant’s advisers. She was, however, forced to concede that the prejudice to the Commissioners consisted in their being out of their money and in their being faced with a very late appeal; she accepted that there was no evidential prejudice. The objection that the Commissioners were out of their money seemed to us to have very little merit since, as we understood the matter, they had not sought to enforce the second assessment and it was for that reason that they had not received payment. Our allowing a late appeal would have little or no impact in that respect. It appeared to us too that the respondents were in no way taken by surprise by the late appeal; not only the statement of case but also the bundle of documents produced for use at the hearing appeared to have been prepared upon the footing that both assessments were under appeal. It seemed to us that it would be quite unjust to refuse the application and shut the appellant out of an appeal because of an obvious oversight where there was no identifiable substantial prejudice to the respondents. Accordingly we allowed Mr Allen’s application.

4The appellant has two directors, Abdun Noor and his wife, although Mr Noor told us, when he gave evidence, that Mrs Noor takes no active part in the running of the company. Mr Noor has lived in this country since 1985, and has throughout worked in the restaurant trade. For some years before the events with which we are concerned, he had been involved in a restaurant business run from the same premises as those used by the appellant, but that business had ceased and the appellant had begun to trade from 1 July 1996, from which date it had also been registered for VAT. Mr Noor told us that he was an active manager of the business, attending every day unless he was out of the country; he was not often abroad. The restaurant (which served meals to eat in the restaurant as well as take-away meals) was open both at lunchtime and in the evening on every day of the week but except on Sundays, when a buffet lunch was offered, the level of evening trade was far greater than that at lunchtime. In fact, lunchtime trade was so light that on occasion there were no customers at all. We accept that part of Mr Noor’s evidence.

5He also told us that his system for taking orders, both for diners and take-away customers, was to write each order on duplicate pads, one copy of each order going to the kitchen and the other being placed on a spike kept on the counter, near to the till. Those copies were used for calculating the price of each customer’s order. At the end of each day’s trade, Mr Noor would count up the takings and also total the prices shown on the meal bills; the total of the meal bills was entered in a daily takings book and tallied against the cash, cheques and credit card slips in the till. Periodically Mr Noor took the takings book and purchase invoices to the appellant’s accountant, who prepared its VAT returns and other accounting records. He told us that he assumed that the accountant did his job properly, and signed what was put in front of him without making any real check of his own. He was, however, satisfied that the appellant’s VAT returns were correct.

6Included within the bundle of documents was a letter, dated 28 August 1996, written to the appellant by its local VAT office, in which it was instructed to keep its complete till rolls, where a till was in use (as it was in the appellant’s case) as well as details of each transaction—in the case of a restaurant the only practicable method of doing so is to retain the meal bills. Mr Noor told us he had not received that letter, and did not know that it was necessary to keep those documents. He did acknowledge having received a letter of 28 January 1999, written following a control visit made, as we were to learn, as an element of the respondents’ enquiries, in which the officer who had made the visit had observed that till rolls and meal bills were not being retained and in which the instruction was repeated. Mr Noor said he had been keeping his till rolls and meal bills since he received that letter, but he had not done so beforehand.

7It was put to Mr Noor in cross-examination that he had been penalised for dishonest evasion of VAT, some years ago, when he was running a restaurant business in partnership with others. He conceded that he had been penalised, but maintained that the dishonesty had been perpetrated by his partners without his knowledge, and that he had suffered the penalty because he was jointly and severally liable with them. Rather to our surprise, he was unable to tell us how many other partners there had been. Miss Grundy suggested to him that, with that experience behind him, he would have been scrupulously careful about keeping his records. Mr Noor responded that he had not understood the importance of record-keeping.

8We mention at this point that we find that contention quite implausible. If, as he would have us believe, Mr Noor had been penalised for the dishonesty of others while himself quite blameless, we are quite sure he would have taken great care to avoid having any further suspicion fall upon him. Even if it were true that he did not receive the first of the two letters to which we have referred (the letter is correctly addressed and we think it more likely than not, and so find, that he did receive it), his professional advisers would almost certainly have advised him that he should retain his records. The requirement is also spelt out in the various notices which retail traders receive from the Commissioners. We do not accept Mr Noor’s evidence that he did not know that he should retain his till rolls and meal bills; on the contrary, we think he was very well aware of the requirement but chose not to retain them.

9The assessments are based upon test purchases and observations carried out by Customs officers which, we understood, were prompted by their perception that the appellant’s declared takings were lower than might be expected for a restaurant of its size and apparent volume of trade. We had evidence from two Customs officers, Peter Wright and Martin Roberts, both of whom had had some involvement in the observation and test purchase exercise, and we had several pages of hand-written observation records as well as a video tape, whose source was a static camera trained upon the entrance to the appellant’s premises for the entirety of one day’s trade, for part of which we also had a hand-written observation record. We did not watch the whole of the video tape, but saw extracts of it which we were invited to compare with the hand-written record.

10Our first impression was that the video tape had significant limitations. It was in monochrome, and it appeared that the camera had been set to a fixed focal point. The combination of these factors made it difficult to discern some details, and it was of course impossible to identify those entering and leaving by the colour of their clothing, although some discrimination by shade was feasible. It was apparent from the video tape that the weather conditions on the day on which it was taken were not good, and the lighting conditions were sometimes less than ideal. There were intervals when the view of the entrance to the restaurant and the pavement area outside was partly obscured by parked vehicles. However, the tape was shown to us by means of a machine which was capable of playing it both forwards and backwards and in slow motion, even at one frame at a time. By the use of this facility, and with some degree of repetition, it was possible to make what we concluded was a quite accurate analysis of the numbers of people entering and leaving the premises, and of what they were carrying. Despite our initial misgivings, we accept that the video tape is, in fact, a reliable record capable of adequate analysis.

11We heard a good deal of quite detailed evidence about the observations, but we do not think it necessary to repeat that detail here. It is sufficient to record that while we accept that the hand-written records are not absolutely accurate, we are not persuaded that any inaccuracy is greater than is inevitably the consequence of human fallibility. One must bear in mind that, whereas it is possible to play a video tape many times over, there is only one, often fleeting, “live” opportunity to observe a person entering or leaving a restaurant, and to see what he is carrying. We are satisfied that, despite that limitation, the Customs officers involved did their best to observe carefully, and to record their observations accurately. That finding does not dispose, however, of Mr Allen’s principal criticism of the observation records, to which we will turn shortly.

12Mr Noor was, of course, unaware at the time they were made of the observations and was unaware also that some purchases of both take-away meals and of meals eaten in the restaurant had been made by Customs officers in the course of their investigation. On Friday 5 February 1999, after several days’ investigations had been undertaken (they had started in June 1998), officers made an unannounced visit to the premises at the conclusion of the day’s trading. Mr Noor told us that he was quite surprised by the visit, but not surprised in the sense that he felt that Customs had caught him out and he was (as the Commissioners concede) quite cooperative with them. Mr Wright accepted that Mr Noor had not attempted to restrict his access to the premises, although he had confined himself mainly to the counter area, and that he had not discovered any “off record” books or similar items.

13The takings for that day were found to be £2,418.55. That figure was not far short of double the highest amount recorded in the appellant’s takings book for any other Friday. It was also observed that the credit card payments for that day amounted to £922.15, or about 38 per cent of the total, much lower than the average shown in the takings book for the preceding Fridays of 1999, and significantly less than the lowest recorded figure during that period of 49.5 per cent. A similar analysis of the takings book for the months of April, May and June 1998 showed that credit card sales represented on average 66.9 per cent of the takings for Fridays; the lowest declared proportion was 58 per cent. Mr Noor agreed, as indeed his takings book indicated, that the volume of the trade at the restaurant did not vary widely from one week to the next, and was not prone to significant seasonal change.

14Mr Noor accepted that the takings on 5 February 1999 were much higher than those recorded on comparable days. The only explanation he could offer was that there had been a large party, of about 13 (though later agreed to be 17) diners, in the restaurant that evening. The Customs officers had themselves observed a large party, though without being able to determine a precise number, and the respondents acknowledge that this factor may have distorted the takings figure. However, it was, they say, insufficient to account for the whole of the difference between the takings on that day and those recorded on other Fridays. An examination of the available figures shows, to our satisfaction, that that contention is correct; if one eliminates the cost of the large party’s meal, and also the cost of meals eaten by Customs officers on that day, the takings remain significantly higher than on any other day as they are recorded in the takings book.

15There were two further days of observations after the visit on 5 February 1999: on 18 February (when the video tape was obtained) and on 5 June 1999. On 31 March 1999 Mr Noor was interviewed by Mr Wright and by Grahame Thompson, another Customs officer from whom we were to hear brief evidence, in the presence of the appellant’s accountant. Very little turns on the interview, but we need to mention that it was conducted after Mr Noor had been shown the Commissioners’ Notice 730, an indication that he was suspected of dishonest evasion.

16Mr Allen made a number of criticisms of the Commissioners’ analysis of the observation records. First, it was, he said, in many cases impossible to match people recorded going into the restaurant with those recorded coming out, largely, it seemed, because of changes in the teams of observing officers during the course of the day and in part because observations began after the start of the day’s trading, or ceased before its end. We agree with that argument; it was a factor which Mr Roberts, too, took into account when he came to calculate the amount by which, it is contended, the appellant had suppressed its takings. It is a particularly important factor because of the respondents’ assumption that a person who remained in the premises for more than a certain length of time had eaten a meal. A time of 30 minutes was mentioned during the course of the interview of Mr Noor, but both Mr Wright and Mr Roberts told us that they had in fact used 40 minutes as the yardstick. It transpired that for part of his calculation, and despite his recollection, Mr Wright had after all used 30 minutes. Mr Noor’s evidence was that it was not possible to order and eat a meal in as short a time as 40 minutes, and certainly not 30 minutes. He considered 45 minutes was the minimum practicable time, and an hour more realistic.

17The only evidence we heard about the time required for a customer to enter the restaurant, be seated, consider the menu and choose and order a meal, for the meal to be prepared, served and eaten, and for the customer to pay his bill, came from Mr Noor, whom we did not consider to be a satisfactory witness; as will be apparent, we reject much of what he told us. Nevertheless, using our own experience and bearing in mind that (except on Sundays) the preponderance of trade was in the evenings (that is, few of the meals served were eaten by those hurrying to return to work after lunch, though it is possible that evening customers were on their way to a theatre or cinema) we accept that, for most customers, even 40 minutes is short; we think there is some merit in Mr Noor’s contention that an hour is more realistic. On the other hand, as Mr Wright was to point out, it is difficult to imagine why a take-away customer would be in the premises for as much as 40 minutes. Since there would be no need for customers to choose from a menu, or for the food to be individually prepared, at the Sunday buffet we think rather less than an hour might be possible, although even then we think that most customers would take at least 40 minutes.