Examination of Witnesses (Questions 160-179)
MR RICHARD BROADBENT AND MR MIKE WELLS
MONDAY 25 FEBRUARY 2002
160. So plus or minus two per cent is the accuracy, but we are talking about a figure here of £450 million out of £22 billion, which is roughly the same kind of percentage or smaller even, is it not?
(Mr Broadbent) Yes. It is an estimate which we have built up as accurately as possible, but we have to work within the limitations, indeed I would observe that one of the reasons we have to be quite careful about at some cost moving an organisation to change its pattern of activity very quickly is to be very certain we do understand what is happening and allow for some of the inaccuracies in the estimates.
161. What assessments have been carried out between the links between the rate of duty and the rate of fraud? Is there a correlation between the rates of duty in the UK and the rates of fraud?
(Mr Broadbent) May I just make one point? It is really to do with the price differential rather than the rate of duty. I say that because, particularly in the case of the Republic of Ireland, VAT rates are different, they are higher, there are also exchange rate differentials which change all the time. The biggest driver of behaviour, certainly in the limited experience we have had using these methodologies, including fraudulent behaviour, is total price which of course materially reflects the oil price. One of the reasons we may have seen a sharp increase last year was that the oil price peaked last year.
162. It will have peaked in Ireland as well, will it not?
(Mr Broadbent) Yes; sorry; I was just mentioning two separate points. In Ireland the differential at pump price is slightly less than you suggest when looking at duty, although it is still large, but the differential also varies with things like exchange rates. For example, the euro has weakened against sterling. In GB the level of fraudulent activity does seem to be related not just to the price of petrol, which reflects amongst other things the oil price as well as duty rates.
163. In terms of your advice to Ministers, in terms of the rate of duty which is levied, what kind of advice are you giving in terms of assessment of the correlation between the rates of duty going up and the levels of fraud going up? You cannot do anything about the price of oil, so that is almost irrelevant, but you can do something about the rate of duty. What assessment are you carrying out in terms of advice to Ministers in terms of the level of fraud correlating with the rates of duty?
(Mr Broadbent) In our advice to Ministers we do try to take account of all factors, including oil price for example. We certainly try to take into account factors which might affect fraudulent demand. We give them the best advice we can on the basis of our best assessment.
164. That is a non-answer really, is it not? What I want to know is what assessments you have done which correlate the rates of duty with the rates of fraud. For example, the rate of duty in 1992 on unleaded petrol was 23 pence per litre and today it is 45.8 pence. What fraud increases have there been between 1992 and the year 2002? Have you done that kind of assessment?
(Mr Broadbent) Yes, we have. In the petrol market the level of fraud is virtually unchanged but the level of fraud in the diesel market has gone up and that is why I am worried about the diesel market. To go to the question you are quite rightly asking me, there is no doubt that fraudulent activity is linked probably to two basic things: one is the price of the product and in this case duty is a large part of the price though not the only thing. The other one seems to be a set of externalities. For example, over the last decade we have witnessed a big increase in the internationalisation of transport, so there are many hauliers who are more aware of differentials in diesel prices. We have witnessed a fuel crisis which may have changed behaviour. We have witnessed different economic drivers which may drive different competitive activity. Fraudulent activity does seem to me to have these two components. I do not want to avoid the duty point, but the oil price is very important as well in terms of driving behaviour, very important, indeed, so are exchange rates in terms of international comparisons. Linked to that there have to be some sort of externalities which sensitise it.
165. How do you disaggregate the oil price issue from the duty price increase?
(Mr Broadbent) We tend to work in terms of prices rather than duty. We had to recognise that duty is a driver of price, but the elasticity work we do is related to price rather than duty.
166. How do you disaggregate the issue of the duty? You say that it is a lot to do with the price of oil. How do you disaggregate the factor of the duty on the oil price when you are doing your adjustment of the overall price?
(Mr Broadbent) We certainly take into account the impact of a duty rate on price and hence on the extent of fraudulent activity. It does also link to the opportunity. It is interesting that in petrol we have seen very little change of fraudulent activity and in diesel there is a change, but a change which dates from a point long after the fuel duty rate escalator was introduced.
167. How much fraud is there in petrol? It is not a huge amount at all.
(Mr Broadbent) About one per cent.
168. You said it was very, very small in terms of the UK petrol. One would expect there to be no difference because there is no fraud, is there?
(Mr Broadbent) The point I was trying to make was that fraud relates not just to price but it relates to opportunity as well. It is not correct that it is only the price, because in fact the trend of fraudulent activity in diesel in GB did not begin as duty rates started to go up. There was a distinct lag. It is clear that some externalities were also introduced into the situation which may have impacted on behaviours.
169. How do you know that? You have just said you do not disaggregate the duty issue from the overall price issue.
(Mr Broadbent) I am not sure I fully follow your point.
170. I am trying to find out how rigorous your intellectual policymaking is when you say that you assess these things on the basis of overall price, albeit duty is a factor in determining that price, but you do not actually disaggregate in terms of your assessment of fraud. I do not know how you do that.
(Mr Broadbent) I hope I am following you. When we look at duty rates we do look at potential impact a changed level of duty would have on demand and revenues.
171. Right; that was the answer I was hoping for.
(Mr Broadbent) We go through the price and I may have misunderstood but that is the way you have to do it technically.
172. What is your department's assessment of the effect on fraud levels of say a 10 per cent cut in duty on diesel? What effect would that have on fraud?
(Mr Broadbent) We find it very difficult to answer that question. We can do some work on the elasticities of demand and we may believe that some of the elasticities we observe are due to fraud, but we do not know enough. You are asking me questions where our ability to work the data which is still very new is quite limited. We do not know what impact a given duty rate change would have on fraud. In some of our regimes we do have and we have done quite a lot of work on elasticities in relation to final price, which is not quite the question you asked me, and you might assume some of the elasticities you have seen in change in demand level were due to changes in pattern of fraud, but I have no methodology at present which
173. Do you not think you should have? For example, the level of tax was a big issue in the 1970s and 1980s. There is a correlation. A lot of academics, particularly in America, have done a lot of work on the correlation between the level of income tax rates and the level of fraud in income tax and they were able to produce academic studies which correlated the two. Why is it that you have not been able to do that in terms of duties on diesel? Is it because you are not rigorous enough, that you have not instructed the right academics, the right economists or do you think it cannot be done, in which case I shall put that to the academics after this meeting is over?
(Mr Broadbent) It is quite a complex problem. First of all, this is all quite new work. Secondly, we do do quite a lot of work on elasticities and a lot of that work is shared with academics. Indeed when we published our estimates in the PBR we published a complete technical paper and invited academics to come and talk about our estimates. We want people to engage in debate with us. There is a full methodology out there.
174. You can instruct them and pay them, but you cannot expect them to do rigorous work on that basis.
(Mr Broadbent) You have to be very clear you set the right question and you cannot answer it yourself before you start paying for people outside. I do not disagree. We do commission work from outside. This may be my fault but we do do a fair amount of work on elasticities and we do try to understand the impact duty rates have on demand and within that patterns of fraud.
175. You could not answer my question about what effect a ten per cent cut in duty would have. It is a very straightforward question if you have done some work.
(Mr Broadbent) It is not a very straightforward question, it is a very complex question. The linkages are very, very complex.
176. Would fraud go up or would fraud go down is my final question?
(Mr Broadbent) I could not answer that question. It would depend quite a lot on the
Mr Gibb: I despair. Thank you.
177. When you get to this stage in the meeting virtually everything has been asked, so I do apologise if I go over some of the points which have been touched upon. However, some of the answers which have been given, have frankly not persuaded me that enough is being done. Mr Jenkins wanted to bring the private sector in and I see it rather differently. Page 8, paragraph 1.10, tells us that 300 staff are wholly dedicated to collecting and protecting the revenue from hydrocarbon oils at a cost of about £10 million. That is a very small number of people involved, is it not?
(Mr Broadbent) It is not the totality of the effort, that is just the totality of the people for whom it is a full-time effort. As the report tries to go on to make clear, the knack with this is to balance the people you have who are absolutely dedicated and full time, because it is a very expensive structural thing to put people in place because you are going to have to move them frequently, you have to have an organisation around them which has the capabilities and the flexibility to move itself to a problem, in this case oils fraud, to add the intelligence input, the investigation input, the legal input.
178. How do you decide what their priorities are because there are several taxes, several products? How do you decide the priorities? What do they concentrate on? What is their main priority?
(Mr Broadbent) That is one of the more challenging tasks for the organisation, but the way we do it is to have an annual process for resourcing the organisation, that process looks at the targets we are set by government, it looks at where we perceive the pressures to be in the world, it looks at what we need to invest to develop future capabilities, which is a very important part of our activity, and how we should allocate the current capabilities to different threats. I should just like to emphasise that for me an important part of this process is that clearly an important judgement is how many you allocate to this task or that task at a point in time. It almost comes back to one of the questions I was asked by the Chairman. One of the things I believe we need to get better at is creating an organisation which is highly flexible, it has sets of capabilities which can be moved flexibly quite quickly, within a year never mind from one year to the next, to tackle problems as they arise because the modern world throws up problems much more quickly than it used to do.
179. Why do you need to be so flexible? We are talking here ... I am not exactly sure what we are talking about. I think we are talking about £1 billion of fraud. We do not seem to be certain exactly what it is, but we are talking of something in the region of £1 billion and you are talking about having to have a flexible staff who can move around. My point is that it costs £10 million to employ 300 staff. Why do you not treble it at a cost of £20 or £30 million? Presumably by doing that, you would get a huge amount more revenue in. It has to be cost effective. You are not speculating to accumulate, are you?
(Mr Broadbent) I should like to say it was that easy, but it is not. Fraud, particularly fraud by organised criminals and smuggling, are amongst the most difficult illegal commercial activities to tackle. Leaving aside the fact that it is quite right for me to be asked to justify my use of existing resources before I am given more, the problems I have are first of all working out what we are going to do to tackle the fact that suddenly we have laundering plants in the UK mainland when two years ago we had none? It is a completely new problem. What are we going to do to solve it? Secondly, assuming I can find a way of doing it, almost certainly that will hinge not just on having more people, it will hinge on having a set of capabilities and skills which unless I have created those capabilities and skills perhaps some time in the past