Piero's Restaurant and Pizzeria (a firm) v Customs and Excise [2004] UKVAT V18570 (22 April 2004)

BAILII is celebrating 24 years of free online access to the law! Would you consider making a contribution?
No donation is too small. If every visitor before 31 December gives just Â£1, it will have a significant impact on BAILII's ability to continue providing free access to the law.
Thank you very much for your support!

[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]

United Kingdom VAT & Duties Tribunals Decisions

You are here: BAILII >> Databases >> United Kingdom VAT & Duties Tribunals Decisions >> Piero's Restaurant and Pizzeria (a firm) v Customs and Excise [2004] UKVAT V18570 (22 April 2004)
URL: http://www.bailii.org/uk/cases/UKVAT/2004/V18570.html
Cite as: [2004] UKVAT V18570

[New search] [Printable RTF version] [Help]

Piero's Restaurant and Pizzeria (a firm) v Customs and Excise [2004] UK V18570 (22 April 2004)

VAT – Assessments – Output tax – Underdeclaration – Best judgment – Restaurant meals – Observations by successive officers eating test meals on 3 days – Average spend per meal for lunch and evening calculated for 3 weeks in 1998 – Applied to number of diners observed on 3 days – Comparison with sales declared for those days – Profitability of business constant over 3 years – Suppression rate applied to 3 years – Assessments reduced – Appeal allowed in part

LONDON TRIBUNAL CENTRE

PIERO'S RESTAURANT AND PIZZERIA (a firm) Appellant

and –

THE COMMISSIONERS OF CUSTOMS AND EXCISE Respondents

Tribunal: THEODORE WALLACE (Chairman)

JOHN ROBINSON

Sitting in public in London on 23-27 February and 1-5 March 2004

Peter Smallwood, of Haines Watts, chartered accountants, for the Appellant

Shaheen Rahman, counsel, instructed by the Solicitor for the Customs and Excise, for the Respondents

© CROWN COPYRIGHT 2004

DECISION

This appeal concerns assessments on a partnership which operates a restaurant in Truro. The assessments were originally made in or before September 1999 for the periods from 1 April 1991 to 30 April 1999 and totalled £252,285 being based on dishonest evasion. By March 2001 the Commissioners had decided not to pursue the evasion allegation, however this was not formally notified until 26 July 2001 when the periods assessed were confined to 07/96 to 04/99, the VAT totalling £110,102 with interest only assessed up to 20 September 1999. A formal notice of assessment for £110,102 plus £25,603.46 interest was issued on 1 August 2001 and the Appellant appealed.

On 28 November 2001 the Commissioners notified the Appellant that the assessment for period 07/96 was withdrawn as being out of time reducing the assessments to £102,485. These were on the basis that 37.51 per cent of output tax in each period had not been declared. The Statement of Case pleaded that the returns "were incorrect in that output tax had been understated".

On 28 April 2003 the Appellant was notified of a reduction in the assessments to £70,042 on the basis that 29.09 per cent of VAT had not been declared. The assessments now under appeal cover the eleven periods from 10/96 to 04/99.

In Decision No.17711 released on 15 July 2002 the President, Stephen Oliver QC, decided against a preliminary submission by the Appellant that the assessments were out of time, holding that they were made on or before 20 September 1999 although no formal Notice of Assessment was issued until 2001.

Preliminary submission

At the outset of the hearing before us Mr Smallwood submitted that the original assessments in 1999 were based on allegations of dishonesty and that although not pleaded in the Statement of Case the rationale of the assessments remained the same. He distinguished the decision in Halil v Customs and Excise Commissioners [1992] VATTR 432 on the basis that there the allegation of dishonesty was implicit whereas here it was the foundation of the original assessment. He accepted that the burden of proof in an assessment appeal is on the Appellant but said that the burden of proof of dishonesty was on Customs who should go first. He cited Deeds Ltd v Commissioners of Customs and Excise (1983) Decision No.1500 and the Human Rights Act 1998.

Miss Rahman said that Customs were not proceeding on the basis of dishonesty and the burden of proof remained on the Appellant. She submitted that Halil was correctly decided and applied here.

The Tribunal accepted Miss Rahman's submissions. In Halil the Tribunal applied the decision of the Court of Appeal in Grunwick Processing Laboratories v Customs and Excise Commissioners [1987] STC 357 which is binding on us.

Assessments outlined

Before considering the evidence we outline the basis of the amended assessments before us. On Friday 13 March, Wednesday 27 May and Monday 27 July 1998 teams of officers ate meals and observed the numbers of diners throughout the time the restaurant was open. They counted a total of 260 at lunch and 304 in the evening; 6 per cent was deducted for children and 4 per cent was deducted for adults not eating meals, the adjusted figures being rounded down. On the basis of three separate weeks, 27 May to 2 June, 13 to 19 June and 1 to 7 July 1998 Customs calculated the average bill per diner at lunch at £9.07 and the average in the evening at £13.02 from bills in the Appellant's records. Combining the numbers observed with the average prices they calculated the projected takings on the three days observed. They did not have bills for March 1998. When applying the average prices to the March numbers they deducted 4.39 per cent to compensate for a price increase. They compared this total with the totals shown on the till rolls for those days after subtracting from the till roll figures the sales which the Appellant had zero-rated as takeaway; this was on the footing that the VAT declared was based on the till rolls. The resultant figure for VAT declared for the three observed days was 29.09 per cent less than the VAT due on Customs' projected takings. This percentage was used to make assessments for all periods from that starting 1 August 1996 to that ending 30 April 1999.

Witnesses

There were two witnesses for the Appellant and eleven for Customs. The Appellant's witnesses were Piero Marci, director of Piero Marci Ltd, which was one of the partners from 10 June 1998, and who was himself a partner up to that date, and Mrs Maggie Fisher, who was a director in the other partner, Evered Stevens Maggi Fisher Ltd. Mr Marci worked at the restaurant in the kitchen full-time. Mrs Fisher was involved in another company with four retail shops which she ran from an office at her home in Newlyn; her involvement was essentially with the figures and finances.

The witnesses for Customs were: Richard John Phillips, who actually made the assessments and organised the test eats and observations as well as taking part on 13 November 1997 and Friday 13 March 1998 and also carried out visits and interviews; Norman Kirton, surveyor, manager of the Truro office, who took part in test eats on 27 May and 27 July 1998; Amanda Hadaway, Rachel Hancock, George Monty Bagwell, Sheila Newman, Peta Andrew and Melanie Tremain, all of whom took part on all three days; David Wilbourne, who took part on 13 March and also 13 November 1997; John Adkins; who took part on 13 March and 27 May; and Alan Beresford, who took part on 13 March and 27 July.

Documents

The documentary evidence was extensive and included the VAT returns; till rolls for 13 March, 27 May and 27 July; Daily Cash Up sheets for the same days; Order bills with till receipts and add lists for 27 May and 27 July; takeaway books; briefing notes for officers for test eats on the three days, notes or jottings made at the time, notes made soon after meals and test eat certificates; visit report on 16 June 1992; notes of telephone conversations, meetings and interviews; correspondence and schedules.

The restaurant

The restaurant was extended in the early part of 1998 to form a new entrance and waiting area. A number of plans were produced, including two scale plans before the work. 72 colour photographs taken last year by the Appellant were put in evidence. There were also a number of sketch plans not to scale made by witnesses. The room was approximately 15 feet wide and 66 feet long with an alley on the right looking up from Kenwyn Street. Before the work this extended the whole length of the room with the entrance on the side of the room about 37 feet up the alley. Beyond the old entrance there was a mild dog leg to the left.

The extension covered the far end of the alley from Kenwyn Street with the door at the top of the alley. The room thus became an 'L' shape, with the main entrance in the base at the far end. Also at the far end was a door leading through to the Mounts Bay Trading Company shop, the entrance to which was in River Street, where there was also a sign for Piero's. Beside the door to the shop were the toilets. Also at that end but facing the new extension was the door to the kitchen. There was a counter outside the kitchen for the pizzeria and then the bar which was quite long. The counter was at a slight angle to the bar. Next to the bar was a table for staff. After this was a raised balcony surrounded by a balustrade with steps leading up to it. At each end of the balcony was a screen with a semi-circular top, wood frames and glazing. The screens were about 10 feet high and were attached to pillars projecting 1' 3" to 1' 6" from the wall; the screens projected about 6 feet into the room. There were shorter screens of the same type on the alley side opposite to the larger screens; those projected about 3 feet. When the extension was built, the screen opposite that nearest to the bar was moved further towards the far end. While the glazing which was coloured interfered with the view, it was possible to see through the glass to some extent.

Piero Marci's evidence

Piero Marci said that he had overall responsibility for catering and the day-to-day running of the restaurant. He said that screens in photographs taken last year were the same as in 1998. He said that waiters took customers to their tables as a table became available. At busy times customers waited in the bar area or, after Easter 1998, in the waiting area opposite the bar. He said that between 1996 and 1999 there had been five or six different head waiters, including his brother-in-law, George. He said that takeaway meals were charged at the same price as meals eaten in the restaurant; takeaway orders were written in a book. After a meal he said that customers usually paid at their table.

He said that there had been no change in accounting procedures since the periods covered by the assessments. There were pads with four sheets white, yellow, green and blue: yellow for the kitchen, green for the pizzeria, blue for the bar and white for the table. The order was written on the pad but not the cost. When the customer asked to pay the blue and white bills were added on an adding machine. The bills were not numbered. Payments were rung into the till. About 30 per cent of payments were in cash. If payment was by credit card, the customer might add a tip. Cheque and credit card payments were rung onto the till. Cashing up was done after lunch and at night. He himself was normally working in the kitchen and only used the till once a week. He was not involved with tips.

Cross-examined, Mr Marci told Miss Rahman that he did not suggest to staff that they should take cash from the till to make up for a tip on a credit card; there was no system, but they had wondered if that happened. If he was there he checked whether the till had been cashed up but did not check the figures or whether it had been done properly. He had done the cashing up at first but then left it to someone else. The 30 per cent cash figure varied quite a lot. He said that staff might ring up a "No sale" to give change for cigarettes, to get a telephone number out of the till, because a customer was going to pay on another day or because of petty cash payments. He said that he could not explain why officers' bills might have been rung in as "No sales".

He said that 40 per cent of customers at lunch came in through the adjoining shop of Mounts Bay Trading Company; River Street was busier and Piero's was advertised outside it. Customers might come through the shop to dine or to buy a takeaway. He said that they bought pizza boxes in bulk and might serve takeaways in foil or even two in a box on occasion. He said that what officers could see depended on where they sat; the screens obscured the view.

Re-examined, he said that only one member of staff remained from 1998. Mrs Fisher had always done the paperwork.

Mr Marci told the Tribunal that he had to trust staff and rely on them. The kitchen bills were not checked against the white bills. The business had been doing well with profits going up every year. Mrs Fisher came in once a week or fortnight. After cashing up, the takings were put into a bag with cash, coin, cheques, cards, cashing up sheets, Z readings and till rolls and placed in a safe place for collection; the bills or add lists were not put in the bag. He was informed if the Z reading disagreed with the cash reconciliation by more than say £5. If the till was short, he might make it up from tips. Petty cash expenses were paid for from the till with the slips being kept. The Z reading was usually attached.

Mrs Fisher's evidence

Mrs Fisher said that she had come to England from Canada in 1946 as a student and had stayed. She had worked in the theatre then moved to Cornwall making pottery. She had been retailing at Mount Bay Trading Co, Penzance since 1978 when she formed the company. She bought the shop at Truro which was originally a hotel with land on which she built a restaurant. Her son was the architect. She first ran a brasserie but this was not successful. She met Piero Marci and suggested a partnership. After a three month trial period he bought 50 per cent of the restaurant business. Her four shops were separately registered for VAT.

She said that she managed the shops and the outside activities of Piero's from an office attached to her home at Newlyn. Anthony Williams & Co had been their accountants since about 1992. They did the purchase ledgers with the VAT returns and payroll for PAYE. She maintained the sales figures, reconciliations and banking figures. The cashing up sheets showed cash, cheques, credit card receipts and petty cash payments. The Z readings were stapled to the cashing up sheets. She banked the cash and cheques.

Mrs Fisher said that before 1998 VAT visits had been congenial. None of the officers had ever queried the procedures until Mr Phillips. The fact that till rolls and journal rolls had not been kept was never an issue. Customs were happy with the records kept on the computer Sage system. The restaurant rolls were kept at the restaurant. Since Mr Phillips' requirement the journal rolls had been kept.

She said that the bags were sealed and dated and kept in the kitchen with the full knowledge of the staff as to their whereabouts. In 1998 Mr Stevens had opened the bags and reported any discrepancies with the Z readings to her. On a few occasions she asked Piero for an explanation for a shortage. If there was no reasonable explanation it was made up from tips. She assumed there would be minor mistakes but she was not casual about discrepancies. A banking reconciliation sheet with gross sales, petty cash payments and net amounts to be banked was provided to the accountants monthly. This was used for VAT output figures. She said that Piero telephoned the figures from the takeaway book to her and those were passed to the accountants to deduct zero-rated sales from the total takings. The VAT return was produced with prior year comparisons to show any anomalies. January to March was typically quiet. Easter was a good period as were school holidays and half terms and the run up to Christmas. August was a good month but September and October were quiet. She kept a close watch on the food purchase figures and also the alcohol figures.

She said that the restaurant was closed on Sundays and holidays. She said that the restaurant could be in Southern Italy. It was a family restaurant staffed by people from Piero's village in Sardinia. All the staff were family or friends. They came speaking little or no English. It was busy and popular and quite different from pizza chains. She said that in 1997 the BBC made a feature film about Piero's and repeated it in 1999; this noticeably affected turnover figures. It was part of a series about West Country eating places. There was no music. The entertainment was the ambiance of staff and customers. They did not take bookings for parties of less than eight.

Mrs Fisher said that she spent a lot of time at the restaurant in March and April 1998 because she was decorating the extension personally. The restaurant stayed open throughout and customers would look behind the curtain to see her painting.

Mrs Fisher said that the restaurant had been designed by her as accessory to the shop. Customers could take a cappuccino into the shop or they could shop after eating. The shop was open to 5.30pm on weekdays with late opening on Thursdays, but was closed on Sundays. The alley and restaurant and shop were used as a way through from the multi-storey car park off Kenwyn Street to River Street. This was not detrimental because it brought custom. She said that she designed the restaurant with her son as architect. She bought the screens many years ago and glazed them herself.

She estimated that pizzas were only 40 per cent of takeaway sales. There was no standard method of packaging although some 1400 pizza boxes had been bought in 1998. Customers might use their own plates, or foil containers or grease proof paper might be used. There was no real difference between eating in and takeaway customers although perhaps more males and children ate in the restaurant. It was very unusual to have large parties in the summer.

The observations

On 13 November 1997 (a Thursday) Mr Phillips and David Wilbourne visited the restaurant to observe the level of trade and to eat a meal. When they arrived at 1pm they counted 40 diners including themselves and during the hour when they were there 12 more diners came in. They paid by cash which was rung into the till. They did not see any takeaway meals.

The Appellant's VAT returns indicated zero-rated sales as being 15 to 25 per cent of gross sales. Mr Phillips decided to undertake a detailed examination of the business. On 19 February 1998 he telephoned Mrs Fisher to make an appointment to see the business records. Mrs Fisher said that the daily cash sheets and till Z readings were not kept once they had been entered onto the computer programme. Mr Phillips told her that the prime records must be kept. He confirmed in a letter dated 4 March 1998 that the till rolls with Z readings and daily cash sheets must be kept, together with the order notes or bills and the takeaway book.

Mr Phillips then prepared a series of covert purchases with teams of officers eating meals over the whole time the restaurant was open on a series of days the first of which was Friday, 13 March. He prepared a brief which was given to all participating officers. This stated that the main objective was to verify whether there was any takeaway trade. It stated that other information was useful including the number of staff, the number of diners and whether a till was being used. The first entry on each shift were to count the number of diners and added parties with later shifts only counting new arrivals. Officers were asked to complete a test eat certificate. No instructions were given as to making notes on site or the retention of such notes, nor was any advice given.

The first shift was covered by Rachel Hancock and Amanda Hadaway, both of whom were based at Exeter. They arrived just before noon and had a coffee which was rung in to the till. A group of 15 people came in who had booked a table and were seated in a long table beside the alley beyond the screens. The officers were given a table between the screens; Mrs Hadaway sat with her back to the alley and Miss Hancock sat facing her. They decide to split the observations, with Miss Hancock taking the area between the screens including the balcony area and Miss Hancock taking the main restaurant area on her left which the large party was sitting and the bar area on her right. Mrs Hadaway made brief notes on a DER Direct leaflet. Miss Hancock made notes on a Kays towels leaflet and also on the DER leaflet.

Miss Hancock noted 26 people including themselves eating on the 9 tables which she was covering, two of which were occupied twice. She was able to record a considerable amount of detail including many of the drinks served to the tables. Mrs Hadaway had a more difficult task since her tables were not so close and were on the other side of the screens. Miss Hancock had noted that three people jointed the party of 15 and overheard someone saying that it seemed that the whole of Cornwall County Council was there. Mrs Hadaway noted on the DER leaflet "5 x 2 2 x 3 1 x 18" and "By bar 1 x 2 1 x 3". After the meal they returned to the office and wrote up their notes, finishing at around 3.30pm. Mrs Hadaway's note included the following, "Between noon and 13.30 I observed 6 tables of 2 3 tables of 3 1 table of 4 and 1 table of 18". We could find no sign of the table of 4 on either leaflet. Given that there was a party of 18 in the restaurant end and that there were two tables by the bar, Mrs Hadaway's note would mean that there were nine more tables occupied in the restaurant end. Unlike Miss Hancock, she made no record of tables being occupied twice over. We find it difficult to see how the extra table of four on her note would have been accommodated. We prefer her contemporaneous note to her later note. We conclude that 65 people took meals during the first shift. Four customers were turned away because the restaurant was full.

The next pair of officers observed 12 additional customers including themselves, two were children. Mr Smallwood accepted this figure. This makes a total of 77 for the luncheon period. This is nine fewer than the figure of 84 used for the assessment. The Z reading on the till roll for this period was £349.50. This was clearly a substantial understatement, even if as Mr Smallwood suggested the party of 16 may have paid on a later day; there was in fact no evidence to support this suggestion.

The first officers to eat in the evening were John Adkins and Alan Beresford. They arrived at 6.30pm when there were six people. It was Mr Beresford's first test eat and he had no training. He counted 59 people sitting down and placing orders, noting them on a newspaper which he said he gave to Mr Adkins but which has been lost. He told us that he used a five bar gate method, with vertical strokes and a diagonal. He said that a count was triggered when a waiter went to a table and took an order. This might have been for a coffee only; he told the Tribunal, "You could be a diner and just have a coffee".. Mr Adkins took no note at the time. He made a very cursory note afterwards, which was not easy to read; the original has been lost. His note said, "3 parties of 2 on entry. Sheila and Tracey 19.10. 19 parties (about). No takeaways. 59 persons entered (… us) Left 19.40 Paid £20.35 cash entered in till. 4 waiters. White stapled together …" He then gave brief details of what they ate. When asked about the number of parties, Mr Adkins said that if it was not in his notes he could not remember. Not surprisingly in view of the lapse of time, he gave a similar answer to most questions. He said that they had no problem seeing people coming in and ordering and that they were surprised the restaurant filled up quickly. Their test certificate dated three days later stated that a number of people joined existing parties. Mr Adkins said in evidence that they stopped counting when Mrs Newman came in at 7.10pm and left at 7.40pm. The assessment was calculated on the basis that 67 dining customers came in during the first shift, comprising the six there at the start, the 59 noted and the two officers.

It is most unsatisfactory that the notes and evidence for this shift were so poor since the number of customers was sharply higher than those observed on the first shift in May and July when the first shift was longer. We return to this after considering the following shifts.

Sheila Newman and Tracey Wheeler recorded their arrival as at 7.12pm. Their record keeping was altogether better. They made notes on a holiday brochure which Miss Newman produced, noting the numbers and times in and out and whether there was a child or baby. The first three notes were "1933 2 O 1940 2 I blonde 1942 2 O (ours)". 'O' meant out and 'I' meant in. From those notes Miss Newman produced a detailed log which they both signed at 10.50pm back at her home which was 50 miles away. Miss Newman did not recall waiting long to be served, however the restaurant was busy.

Between 7.33pm and 8.01pm they recorded 20 customers coming in and a woman and child who bought a takeaway and six customers including the two officers leaving. At 8.03pm four customers came in and were told that they would have to wait 20 minutes. The restaurant was therefore full at 8.03pm. Between then and 8.30pm when the next shift arrived they counted another 23 customers in and 24 customers out, excluding a man who came in and went out at 8.27pm and the next two officers. Nine females were logged in at 7.45pm and eight were logged out at 8.07pm; Miss Newman said that the fact of nine females coming in stuck out but she agreed with Mr Smallwood that the eight going out might have been the same party. They made a sketch plan which showed five tables for two or four plus tables for eight and 16 at the restaurant on Kenwyn Street and, seven tables with seating for 16 in the balcony area, table for nine and two where they were opposite the balcony, two tables for two opposite the bar and a table for seven between the balcony and the bar. Miss Newman went to the toilet twice to collect her thoughts and to get a chance to see more. Most of the notes on the brochure were by her colleague. Some of those who came in joined parties already there. She could not say whether she saw the customers who were told that they had to wait 20 minutes eating.

Peta Andrew and Melanie Tremain who arrived at 8.30pm were standing near the bar for 15 minutes before being shown to the table where Miss Newman had been sitting. The restaurant was very busy when they arrived. They had a glass of wine each when waiting; this was rung up. They kept no notes at the time. They counted the number of people being shown to tables and kept checking with each other. Mrs Tremain could not say whether everyone who sat down dined. She had done similar operations before but had received no training. They left at 10.10pm and Miss Andrew made brief notes in a car outside. She noted among other things, "19 customers 2 x 2 2 x 4 1 x 5 1 x 2 then RP and DW arrived 11 staff included kitchen staff No takeaway".

Mr Phillips and Mr Wilbourne arrived at 10.00pm and were shown to a table. Nearly all tables were occupied. Mr Phillips made brief notes on a napkin and wrote up a test certificate back at the Truro VAT office that evening. Apart from themselves the only customers to come in were a party of a man and four girls. When they arrived there was a party of girls next to their table. They saw no takeaways.

Mr Phillips was told on a later occasion that the maximum seating capacity was 72 and did not dispute this figure. Since the tables were of varying shapes, some being round, some square and some oblong and the tables were not fixed, the number of seats at a table could be varied by putting tables together and moving chairs. Miss Newman's diagram showed a table of 16 where Mrs Hadaway had noted 18 at lunchtime. Miss Newman's diagram must have shown the numbers on tables at different times since it totalled 76 seats. In our judgment it will be rare for a restaurant to get each seat occupied. Many customers would object to being obliged to share tables. Inevitably there will be tables for four with only two customers. We would estimate the practical maximum at any one time to have been 60; on 27 May customers were turned away when a maximum of 58 people were there.

On this basis we do not find the numbers of apparent diners noted on the first evening shift to be credible. Miss Newman's careful note recorded 20 people coming in from 7.33pm and six parties of two leaving by 8.30pm at which stage all tables must have been occupied. From this it follows that there must have been room for 14 more customers at 7.33pm. Mr Adkins and Mr Beresford were only observing for an hour. It is reasonable to assume that the six there when they arrived may have left by 7.30pm. The till roll shows £23 at 6.58pm and four payments between £5.30 and £6 up to 7.30pm as well as three no sales. It also shows five rings of £3.90 or less. If six customers left before 7.30pm then it is improbable that more than 52 dining customers came in before 7.30pm, six before the officers arrived and 46 while they were there. Mr Beresford did not in fact distinguish between those merely ordering coffee and those dining. It cannot in fact have been that easy to distinguish them. The till roll for the early part of the evening contains no fewer than nine rings which seem too small to be a meal. We find that there were 52 dining customers during the first shift and that the balance of those seen were not in fact eating.

The question of the eight females seen going out at 8.07pm is the next issue. Miss Newman accepted that they might have come in at 7.45pm. Her diagram showed a table with nine next to her table and also showed a second table of eight on the other side of the screen from the balcony. Mr Phillips' test certificate recorded a party of girls to his left. He was sitting next to where Miss Newman showed a tale of nine. We conclude that the party of girls seen by Mr Phillips were the nine females who Miss Newman saw come in at 7.45pm, although that was over two hours earlier.

We find therefore that there were 77 diners at luncheon and 127 in the evening, compared with the 86 and 142 on which the assessment was calculated.

We now turn to 27 May which was a Wednesday, two days after the Spring Bank Holiday. Briefs were issued again on this occasion with substantially the same content. Again there appears to have been no oral briefing as to observing and no de-briefing afterwards.

Miss Newman and Mr Kirton covered the period from noon to 1.30pm. They worked closely together. Mr Kirton recorded customers as they passed going to a table and received a menu. Mr Kirton had a good view; he could see through the glass in the screens and could see the waiters bringing food to the tables. Miss Newman noted customers coming in and going out on a Sovereign holiday brochure together with the times from which they compiled a detailed log. At 1.15pm customers were turned away because the restaurant was too full and five minutes later a couple left because it was too busy. Fifteen of the 82 customers they recorded were children; another four were teenagers. Miss Newman recorded the words "started going out" against two adults and two recorded as coming in at 12.46pm. There were amendments to other notes on the brochure. If they did not go out, it is surprising that that note was not crossed out. We conclude that they did leave and did not eat. Mr Kirton noted that at 1.00pm 22 tables were being used for eating purposes and all were occupied; in view of the plan with the notes, a number of those must have been joined together. There were also easy chairs near the entrance where other customers had coffee or drinks. We conclude that there were 78 customers who ate meals of whom 13 were children and 4 were teenagers.

The second lunchtime shift on 27 May was covered by Mr Adkins and James Holmes who did not give evidence. Mr Holmes kept some notes on a newspaper. Apart from themselves the officers noted 3 tables of two and one of three. We accept their total of 11.

The first evening period to 8.10pm was covered by Miss Andrew and Mrs Tremain. They made no notes at the restaurant. They counted customers being shown to tables by waiters and checked back with each other. They relied on their memories when making a note later that night. The note was brief. The only parts relevant to numbers was "38 people sat down to eat not including us No takeaways … Stopped counting when Amanda and Rachel arrived and sat down at a table." The certificate typed on the next day contained no further detail on this aspect. Miss Andrew recalled seeing children but could not say how many. It is extremely difficult to keep a reliable count of this type in the head. Given that they were counting people in, any error is more likely to have overstated the numbers through double counting. We consider that a total of 37 should be used.

Mrs Hadaway and Miss Hancock arrived at 7.58pm and were given a table opposite the balcony. Miss Hancock's notebook was dated on the following day. Miss Hancock said that they would have made trigger notes at the time and left them with Mr Phillips. These notes have been lost. The notebook contains specific times of arrival for 14 males, 19 females and a child besides themselves. We accept that they did make a note at the time on which the notebook was based and we accept the figure of 36 including themselves.

The last two officers on that day reported two customers dining apart from themselves.

We find that there were 89 customers dining at lunchtime on 27 May and 77 in the evening, compared with 93 and 80 used in calculating the assessment.

We next turn to 27 July, a Monday. Miss Newman and Mr Kirton covered the period to 1.30pm. Again Miss Newman kept a log at the time using a travel brochure, noting the numbers and times of arrival and departure. Again a helpful note was prepared from this. We accept the figure of 57 used for the assessment. When they left it there were 44 customers excluding the next two officers.

Mr Beresford arrived next with Mr Holmes. Their report signed later that day stated that the restaurant was over three-quarters full when they arrived; this agrees with the previous officers' log and our figure at paragraph 40 of a practical capacity of 60. Mr Holmes kept trigger notes on a Halifax loan form. They counted 22 customers other than themselves of whom four were children and two were teenagers. We accept the figure of 24 for this session.

Mr Smallwood did not challenge the figures of 44 and 2 for the first and last evening sessions on that day and we accept them. The period from 8.00pm to 9.30pm was covered by Miss Andrew and Mrs Tremain. Again they took no notes at the time, counting in their heads. Miss Andrew saw waiters serving foods to all tables but six years later could not say that all observed were dining. Mrs Tremain made a note later on that day which both signed. Insofar as relevant, the note said, "30 diners (not us) … No children amongst diners … 4 diners before Monty and Dave left (check with M & Dave). Busy but few empty tables." Our comments at paragraph 47 are equally applicable to the same officers here. Their certificate stated "During the time we were on the premises we noted thirty people (excluding ourselves) who dined." It seems to us that that 30 must have included the four. We discount their number by two to allow for error.

We find that 81 dined at lunchtime on 27 July and 76 in the evening, compared with 81 and 82 used when calculating the assessment.

In reaching the above conclusions we have taken account of Mr Smallwood's submissions as to what the officers could have seen given the presence of the screens. We do not consider that it would have been at all easy to tell whether a customer on the other side of a screen was eating. However if a customer was shown to a table and given a menu and a waiter later brought food to the table, it is a reasonable assumption that the customer was eating a meal. No evidence was given by the Appellant that customers were shown to tables other than those near the bar if they merely wanted to drink and not eat. However there was evidence of customers leaving when the restaurant was busy and of other persons joining existing tables. It is obvious that the more crowded the restaurant was the more difficult it must have been to get a good view of those at the far end.

Adults not eating

The reduction to the assessments in April 2003 were in part to reflect adults entering, joining groups, not eating and not appearing on order bills. A reduction of 4 per cent was made on this account. This reduction was from observed numbers some of which took account of customers who left without eating and some of which did not. Miss Rahman said that the allowance of 4 per cent was high enough to cover any errors appearing from the evidence and that the Tribunal should avoid duplicating allowances for errors. We have considered this submission but have concluded that the allowance of 4 per cent should stand in addition to the adjustments which we have made to the observed numbers. The fact is that the observations were of uneven quality. But even the most careful observations based on customers being shown to tables and served carried the risk of over counting. We have already commented on the difficulty of observing customers at the far end when the restaurant was crowded, so that the view would be hampered both by the screens and by closer customers.

Children

The next issue which arises is the treatment of children. Since the calculation of the assessments depended on multiplying the number of diners by average meal prices, the assessments would be higher if children were included in the multiplier who shared meals or ate and drank substantially less than the average. Soft drinks were not of course free but they were less expensive than alcohol. The briefs gave no instructions as to the treatment of children and some officers did not distinguish between children and adults. Miss Newman was the only officer to distinguish between teenagers and children. A child might be a toddler or might be a teenager. Their appetites could clearly vary widely from those eating as much as an adult to those under seven for whom the Bambini Menu costing £3.20 would be substantial. When making the latest amendment, Customs allowed a reduction of 6 per cent to allow for children being one-half of the 11 per cent of children observed in May and July rounded up. It is clear that the number of children on 27 May was underrecorded but not by how much. However the percentage derived from May and July was applied to 13 March which was during the school term when the number of children must clearly have been much lower. We accept Miss Rahman's submission that no change need be made for the 6 per cent allowance for children.

Average meal prices

This brings us to the meal prices used. The amended assessments were based on the average value shown on the order bills for 3 weeks in 1998, 27 May to 2 June, 13 to 19 June and 1 to 7 July. The total for lunchtime bills for those weeks was £5,805.95; the number of diners shown on the bills was 640 giving an average of £9.07. This included £47.20 for 4 meals by officers on 27 May. If those are excluded the average is £9.05. A similar exercise for the evenings results in a reduction from £13.02 to £13.01. The difference is small however when projected over 11 accounting periods it is around £80. Apart from the inclusion of the officers' meals, Mr Smallwood did not make any substantial criticism of the calculation of average meal values apart from the dates used.

Mr Smallwood submitted that meal values derived by the Appellant from bills in February 2000 should be included to give a more balanced sample. Originally the Appellant had contended that these should be substituted. Their calculations were based on bills for four weeks ending on 26 February 2000. The Appellant calculated that the average spend per head was £8.07 at lunch and £10.45 in the evening. They estimated that drinks represented 25 per cent of the average bill and that there had been a 15 per cent increase in drinks since 1998. Adjusting for this they arrived at an average of £7.81 for lunch in 1998 and £10.11 in the evening. The calculations were exhibited to Mrs Fisher's statement, however they were not her calculations, the bills were not exhibited and there was no evidence as to the conditions under which the material used was obtained.

Miss Rahman said that the resultant figures for food only, £5.86 at lunch and £7.58 in the evening, obtained by deducting 25 per cent for drink was not credible in view of the menu prices in 1998. We accept this as a valid criticism.

Miss Rahman also submitted that there was no reason why the Appellant should not have used order bills from late 1998, which had been kept, rather than bills for a period not covered by the assessments. Mr Smallwood said that the February 2000 bills were used because it was only when the letter of 20 September 1999 was received stating that an assessment had been made that the Appellant had realised the importance of the exercise and had carried out an exercise monitoring the bills. This explanation would have been more convincing if there had been proper evidence as to how the exercise was carried out and what measures were taken to ensure the accuracy of the bills.

In our judgment the calculations by Customs which were based on three weeks before the peak holiday season were not an unsatisfactory sample and had the merit of covering one of the test days and being between the other two. When applying their meal prices to March 1998 the Commissioners applied a reduction of 4.39 per cent to offset a price increase at Easter 1998, a figure which was not in dispute.

Takeaway meals

Mr Smallwood directed a substantial number of questions in cross-examination of the observing officers to the number of takeaway meals observed. The Appellant had zero-rated pizzas and other meals taken away by customers on the basis that freshly-baked or cooked food was not "hot food", presumably relying on John Pimblett & Sons Ltd v Customs and Excise Commissioners [1988] STC 358. Customs do not accept that the principle in that case applied to the Appellant. However that was no part of the assessments which were based solely on customers dining in the restaurant.

Mr Smallwood sought to establish that the officers failed to observe a substantial number of takeaway sales and argued from this that their evidence of customers observed dining was unreliable. The Appellant produced small books measuring 4" by 6" in which orders were listed. The books were grubby as Mrs Fisher had said. They contained orders with prices in different ink and daily totals. It was the Appellant's case that the takeaway sales were rung into the till and the assessments were calculated on this basis. The pages for 13 March contained 19 entries and a total of £135.70. We could find no correlation between the figures in the book and those on the till roll. The Z reading for the lunchtime session on 13 March was £349.50 and that for the evening session was £1,205.95. The takeaway book for 27 May showed 23 entries totalling £193.20. Again there was no apparent correlation with the till roll. The Z readings were £668.60 and £833.60. The takeaway book for 27 July showed 25 entries totalling £224.70 and again we could find no correlation with the figures on the till roll. The Z readings for that day were £500.60 and £940.90.

Mrs Fisher produced bills for 1300 pizza boxes between March 1998 and February 1999, normally at 100 a month. This would suggest up to four being used daily. We note that the takeaway book for 13 March appears to show 13 pizzas. If the number of takeaways in the books for the three days is projected on an annual basis this would give over 8000 takeaways per year. We do not find this to be credible. The officers were looking for takeaways and saw very few. We accept that they might have missed a small number through customers leaving by the shop or on the later days by the new entrance particularly since the takeaway counter was some distance away from where they were sitting. However we do not accept that this undermines their basic credibility as to the number of diners seen. We consider it more probable that they missed some of the takeaways when watching the much larger number of diners than the reverse. The diners passed their tables whereas, apart from 13 March before the entrance was changed, takeaway customers did not do so.

Mr Smallwood also submitted that some of the takeaway customers may have been counted as diners. This seems to us to be improbable since the diners were at tables either close to the officers or in the primary restaurant area towards Kenwyn Street. Whether takeaways were in boxes or bags we do not consider it at all likely that the officers would have missed them. Furthermore although the prices were the same the average takeaway order was much less than those of diners, and it would not have been in the Appellant's interests to seat takeaway customers at tables which would otherwise hold diners. There was no evidence from Mr Marci that takeaway customers were seated in the dining area. We do not accept Mr Smallwood's submission on this aspect.

Mark-up

Mr Smallwood also submitted that the underdeclaration assumed by the assessment involved an addition to sales as shown in the accounts which would give an unrealistic mark-up. He said that the mark-up in the year to 31 January 1998 was 181 per cent which Mr Phillips had accepted as reasonable. He submitted that on the basis of the assessment the mark-up was 296 per cent. We accept Miss Rahman's comment that the cost of materials in this type of restaurant is not high. Furthermore the possibility cannot be ignored that part of the underdeclared sales were applied in buying underdeclared materials.

Best judgment

Mr Smallwood also submitted that the assessments were not to best judgment because when made the assessments extended beyond three years and involved dishonesty. He said that Mr Phillips had said that he had made the assessments in July 1999. The methodology was the same throughout and involved an allegation of dishonesty. Mr Phillips had said that the making of an assessment beyond 3 years and the imposition of evasion penalties were not a matter for him but for someone more senior. Mr Alexander had written on 20 September 1999 that the matter was being reported as being potential civil evasion. Mr Smallwood said that it was irrational to assess on the basis that Customs were considering dishonesty; if the decision was irrational, it was made on a whim and therefore not to best judgment. This submission was apparently based on paragraph 29 of Customs and Excise Commissioners v Pegasus Birds Ltd [2003] STC 262, which is under appeal to the Court of Appeal.

Miss Rahman said that there was no evidence of bad faith by Customs in making the assessments and this had not been alleged. She submitted that the percentage underdeclaration had been calculated by Mr Phillips in good faith and to best judgment.

We have no doubt that the assessments whether actually made by Mr Phillips or Mr Alexander were made in good faith. They followed substantial work and were in no sense irrational or capricious. It does appear that there is confusion among Customs officers as to who has authority to make an assessment for an extended period. However the assessments under appeal were within the normal time limits and clearly could be made by Mr Phillips or Mr Alexander. Like Mr Oliver in decision No.17711, we are troubled by the fact that Mr Alexander notified the assessments in a "without prejudice" letter. We are also unhappy at the concept of Mr Phillips making assessments in July which were not notified until Mr Alexander's letter of 20 September 1999. However those are matters which give rise to criticism of the Commissioners' practice and administration and do not involve the assessments not being to best judgment. It is to be noted that the Commissioners have informed the Appellant that default interest will not be imposed from 20 September 1999. We find that the assessments were made to best judgment.

The period covered by the assessments

Mr Phillips gave the following explanation as to why the assessments based on observations in 1998 were carried back to 1 August 1996. He said that Customs established that on three different days spanning five months there were more customers dining than appeared on the order bills and meals eaten by officers were rung as "No sales". He accepted now that the officers' meals in March may have been declared. He said that Mrs Fisher had said that "No sales" might be to compensate for over-rings or to keep a slush fund to enable the till to be reconciled; he did not regard this as credible. The Appellant had stated that there was no real management of the staff or the operation of the till. It was stated that the till roll and daily cash sheets were reconciled daily and were in close agreement. Evidence of the three days indicated systematic underdeclaration. The profitability of the business was constant over the period in spite of the errors discovered. He concluded that the errors were constant. If the errors had just started in March or May they would have adversely affected the gross profit ratio which would have been picked up by Mrs Fisher.

We accept the validity of this reasoning.

Quantum

We turn finally to the actual quantum of the assessments, applying our findings of fact and using the same methodology as that used by Mr Phillips.

On 13 March we take 77 diners at lunch at £9.05 less 4.39 per cent. Deducting 6 per cent for children and a further 4 per cent for adults not eating, the number at lunch rounded down is 69. 69 diners at £9.05 less 4.39 per cent to offset the price increase results in £597.04 at lunch. For the evening we take 127 diners at £13.01 less 4.39 per cent. Adjusted for children and adults not eating the number is 114 diners. This results in assumed takings of £1,418.03 in the evening and in a total of £2,015.07 for the day.

Contrary to Mr Phillips' understanding, the VAT returns were based on the daily cash sheets which totalled £1556.75, from which £135.70 was deducted based on the takeaway book, giving £1421.05 declared sales or 70.52 per cent of assumed sales for that day.

On 27 May we take 89 customers at lunch at £9.05. Adjusted for children and adults not eating, this gives 80 customers and assumed takings of £724.00. 77 customers in the evening at £13.01, adjusted to 69 customers, gives assumed takings of £897.69. Assumed takings for the day were therefore £1621.69.

The daily cash sheets showed £672.45 and £834.95 and the takeaway book £193.40. The result was £1314.00. That is 81.03 per cent of assumed sales for 27 May.

On 27 July we take 81 at lunch at £9.05. Adjusted this gives 73 customers and assumed takings of £660.65. 76 diners at £13.01 in the evening adjusted to 68 gives £884.68. The assumed total is therefore £1545.33.

The daily cash sheets showed £505.60 and £940.90 and the takeaway book £224.70 giving declared sales of £1221.80 which are 76.06 per cent of assumed sales.

The total assumed sales for the three days are £5182.09 compared with declared sales of £3956.85. This gives 23.64 per cent of sales underdeclared compared with 29.09 per cent used for the reduced assessments. The underdeclared sales were £1225.24 which is 30.965 per cent of declared sales. Applying 30.965 per cent to the declared sales gives the following results rounded down to the nearest pound:

VAT declared adjusted VAT due

10/96 £14,811.04 £4,586

01/97 13,468.14 4,170

04/97 13,026.23 4,033

07/97 14,458.80 4,477

10/97 17,761.12 5,499

01/98 16,627.86 5,148

04/98 15,226.55 4,714

07/98 15,173.15 4,698

98 16,900.56 5,233

01/99 16,371.23 5,069

04/99 16,921.47 5,239

£52,866

The result is that the appeals have succeeded in part, the original appeal having been against assessments of £110,512 which were reduced in 2003 to £70,042; the Appellant failed on the best judgment issue. The Appellant applied for costs if successful. It is likely that any award for costs will be for a proportion only. We direct that any application for costs in the light of our decision be made within 28 days of the release of the decision to the parties.

THEODORE WALLACE

CHAIRMAN

RELEASED:

LON/01/927

BAILII: Copyright Policy | Disclaimers | Privacy Policy | Feedback | Donate to BAILII
URL: http://www.bailii.org/uk/cases/UKVAT/2004/V18570.html