Select Committee on Public Accounts Minutes of Evidence

Examination of Witnesses (Questions 20-39)



  20.  3.6 on page 23. Is 7 per cent of the total tax received, this obviously includes VAT, quite a high figure to be losing?
  (Mr Broadbent) It is quite a high figure and I understand what you are saying. The excise diversion fraud over that whole period of time amounted to a fairly small percentage of the total alcohol revenues. I want to make it clear that it is not a point I labour because these are very large absolute amounts of money. Levels of fraud are too high in my view, this is what the 7 per cent refers to, this is the inward diversion and smuggling coming into play after the outward diversion was dealt with. That 7 per cent, in my view, was too high. What we have been doing over the last nine months is to take a series of operational steps, as I described earlier. I am quite clear we need to do more. We need to produce a strategy for getting that figure down.

  21.  Would you say, table 6 on page 24, if you were the Chancellor of the Exchequer and you were preparing your budget one of the things you might rely on is Customs and Excise providing you with a reasonably accurate estimate of how much money is going to be collected in duties, otherwise it is difficult to set your budget. Table 6 seems to suggest that in every single year in the last decade your initial forecast has overestimated the actual duties received.
  (Mr Broadbent) In the case of spirits the initial receipts did over estimate the forecasts and each year an adjustment was made and the out-turn was quite close to the adjustment, which was made fairly early in the year. That was clearly a sign which was interpreted one way, it was interpreted as a switch from consumption of spirits to wine. That explanation was rather different.

  22.  It is not very helpful for the Chancellor of the Exchequer, is it?
  (Mr Broadbent) The table here has been drawn with quite a large left-hand scale. If you drew it to the scale of our total revenues, this difference between 2.1, or 2.2, these figures would disappear in the balance.

  23.  In 1999-2000 you are, sort of, 600 million out, are you not?
  (Mr Broadbent) In 1999-2000 we are of the order of 500 million below the initial forecast but above the adjusted forecast, so we are between the initial and the adjusted.

  24.  What are you doing to improve your initial forecast?
  (Mr Broadbent) The most important thing I think we should do is to make sure that when we forecast revenues we have a clearer idea about the total amount of tax which is there to be collected as well as estimating the amount we collect. I believe that one of the problems, particularly in these alcohol forecasts, is that if one did not have a very clear view of the total market it was quite difficult to estimate how much revenue you were going to collect. In practice the estimating record within the tolerances which are worked in these forecasts, which I am taking all of our regimes as a whole, is a good one. Within that the estimate for spirits, as you can see, is rather variable. I think it would be less variable if in the estimating process we had spent more time seeking to estimate the total tax available for collection and then worked out how much tax we were going to collect.

  25.  If I can turn to another issue, which is the question of guarantees that you received from warehouse operators. Do you think the guarantee system works well?
  (Mr Broadbent) I think it is working better now. We have tightened up a lot. The guarantee system was loosened somewhat in the mid 1990s as part of a continuing debate within the organisation to balance revenue security with compliance costs. In retrospect that loosening went a bit far and we have now tightened it up again. As a control mechanism its usefulness is material but limited. We cannot rely wholly on guarantees for several reasons. One is that the guarantees are not big enough to cover serious organised fraud. Secondly, because the person who gives the guarantee may not be the person who is at fault. It is always quite a decision to call in a guarantee from an innocent party.

  26.  According to paragraph 1.15 when they were set up they were intended to cover occasional or minor losses and they were set at the level of a single shipment, which I think it says elsewhere in the Report somewhere between 20,000 for beer, 100,000 for spirits. Yet on paragraph 1.19 the most astonishing figure, I have to say, was the 350 million worth of losses which occurred at one warehouse. Would I be right in assuming that that warehouse operator or owner or whoever their guarantee was for 100,000 or 80,000?
  (Mr Broadbent) I do not know what the warehouse guarantee in that case was, it might range up to 1 million. You are right, that was the point I was also trying to get at. To over-rely on a guarantee is foolhardy because you cannot ask for the guarantee to cover the full amount of the revenue.

  27.  These days are the guarantees set at a higher level?
  (Mr Broadbent) The guarantees are set largely in percentage terms relative to the value of the goods moving through, up to about 1 million. 1 million is a large amount of money and I think it is sufficient to do what the guarantee should do, which is to deter the casual law breaker and significantly encourage very rigorous controls by the warehouse keepers. It is not an absolute guarantee against fraud.

  28.  Presumably that deterrence is undermined in many cases when prosecutions do not appear on the horizons? Warehouse operators get an indemnity to cooperate with your officials. Obviously it was a fairly lax regime until recently, 53 per cent of the total amount we are talking about were covered by indemnities, do you still use indemnities?
  (Mr Broadbent) We do still have indemnities to use if we need. Going back to the point I made to the Chairman, the control regime now introduced requires indemnities to be authorised in exactly the same way as letting a load run is authorised. For example, I would be consulted if an indemnity is in excess of 1 million. In fact so far this year no indemnities have been given, which tells you something about how much stricter our controls are.

  29.  Is it fair to say they are no longer your favoured method of getting cooperation?
  (Mr Broadbent) I think it is more that the nature of the fraud has changed. I would have no hesitation using indemnities again if I was faced with a fraud where I needed the collaboration of a warehouse keeper. One thing I have to say is it is a particularly difficult fraud to tackle. This is partly why it took a little time, as you said, to be knocked on the head. It is particularly difficulty because at the point the load leaves the warehouse it is quite legal, you cannot show that the load is illegal until it is gone out of the warehouse. You are faced with a very, very particular problem, which requires a very, very particular way of tackling it. You are asking the warehouse keeper to cooperate so we can follow loads. That is why it took a little bit of time to pick this fraud and that is also why indemnities were used. We are using them less now not because we would not use them in a similar situation but because we do not face that particular fraud.

  30.  Could I ask in the short-term a theoretical question which is not covered in the Report, do you have any analysis if duties are raised, is there a correlation for each pence you increase the duty does the risk of fraud increase a certain percentage or the amount of fraud increase a certain percentage?
  (Mr Broadbent) There is not a calculation that one can do that allows that to happen. Fraud, in our view, is effectively exploitation of economic opportunity. All economic opportunity is judged in relative not absolute terms. The degree of fraud you get depends as much on other opportunities. In broad terms if there is a higher duty it creates a better opportunity relative to other things not changing, but other things are always changing, so it is very hard to know.

  31.  When ministers are preparing the budget arrangements with their Treasury officials is there some kind of analysis from Custom and Excise which says, hold on, if you increase spirit duties by three pence there is going to be an increase in fraud and the returns you are going to get from those increases are not going to be as much as you might expect.
  (Mr Broadbent) When we produce forecasts for the Chancellor on the effect of tax on all our commodities we include in our forecast an offset for expected fall in consumption.

  32.  You must have some method of working out what that might be, some formula or guesstimate, or whatever it is?
  (Mr Broadbent) It is very difficult looking forward to distinguish between the impact on consumption and the impact on fraud. It is very hard to look forward and distinguish these things. We do have some analysis which sets out what will be raised for putting up a duty. We do not within that calculate separately how much is due to fraud, how much is due to personal consumption and how much is due to changing tastes.

Mr Steinberg

  33.  I think this is a very damning report Mr Broadbent, do you?
  (Mr Broadbent) I do not think it is a very damning report. I think it is a very serious report. My reading of the report is it says that there was a change in the world, that created fraud. The report I think says that we reacted quite quickly in terms of law enforcement. It also says that we followed that up by tightening controls. In 1998 we brought it under control.

  34.  You acted very quickly. As a matter of interest, tell me, can you estimate the loss in revenue to the revenue of fraud in alcohol and cigarettes transactions and alcohol and cigarette smuggling per year?
  (Mr Broadbent) We are seeking to make estimates of fraud in both alcohol and—

  35.  I am not asking what you are seeking, can you give me an estimate of how much it is?
  (Mr Broadbent) The estimate for tobacco smuggling was published last year and those figures showed the market in fraud had grown very rapidly over four years, from 3 per cent to 18 per cent. That was projected at a growth of 25 per cent, which is why the government put in place the Tobacco Strategy last year with the aim of, first of all, holding the market down.

  36.  What about alcohol?
  (Mr Broadbent) There are estimates published in the NAO Report which give figures for beer, wine and spirits. Those estimates are based on our own internal estimates and the Government will be publishing full estimates.

  37.  Can the Treasury answer that question?
  (Mr Glicksman) Not to any extent beyond what Mr Broadbent has just said.

  38.  I find it quite amazing that neither of you can tell us the estimate of the amount that is being smuggled in terms of value. I mean it has only been going on now since the single market, 10 years, it is incredible.
  (Mr Broadbent) I did give you those estimates. I will repeat them if you wish. With tobacco the market has grown from 18 per cent—

  39.  I asked you about alcohol.
  (Mr Broadbent) The NAO report contains estimates based on our internal estimates, which will be published shortly. The estimates in the NAO report give a figure for beer, wine and spirits.


previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2002
Prepared 17 May 2002