Examination of Witness (Questions 100
WEDNESDAY 14 NOVEMBER 2001
100. May I ask why?
(Mr Broadbent) Because they are operational matters.
Ministers tend to focus on the strategic matters.
101. Operational matters. That is all right
then. That is fine. Not. Do you think if this were to happen again
Ministers might want to be advised of operational matters?
(Mr Broadbent) I do not think that Ministers particularly
want to be troubled by somebody going to them
102. Ministers do not want to be troubled.
(Mr Broadbent) If you let me finish it would be better.
103. Please finish.
(Mr Broadbent) I do not think Ministers particularly
want to be troubled by somebody going to them frequently saying
they might have a problem here, I may have a problem here, I have
half a problem and am not sure whether it is a complete problem.
They pay us to make judgements about when to go to them and say
"I have made a judgement that I have a problem here. Here
is the size and shape of it. Here are the options". I draw
no distinction about operational matters. I tell Ministers what
I believe is material, what I believe is well founded, what I
judge we know enough about to give them a sensible view and they
are able to take action on. Whether it is operational or strategic
is not to my mind the issue. There is a question about simply
going to Ministers and ringing your hands in front of them, which
I do not think is helpful to them at all.
104. Finally, the Alcohol and Tobacco Fraud
Review. According to Mr Roques this ". . . provided a real
opportunity to make a step change in the Department's response
to the growth in fraud and smuggling of alcohol and tobacco".
His words are, "but this opportunity was not seized".
He talks about "The failure to include estimates of losses
in the report and the slow rate of implementation of recommendations
all contributed to the failure to make such a steep change".
Who is responsible for the failures in relation to that review
that Roques describes?
(Mr Broadbent) In terms of implementing recommendations,
I am not sure he is suggesting that they could necessarily have
been implemented significantly more quickly. Some of them had
been slow to implement because a lot of them involved actions
in the EU. It is actually to the credit of the ATFR and one of
its successes that it did result in a significant raising of the
profile of this issue in the EU with some quite important consequences
coming down the line.
105. What about failure to include estimates
(Mr Broadbent) What Mr Roques does feel strongly about
is that if the ATFR had included estimates, it would have stimulated
debate and led to earlier action. In hindsight that is quite difficult
to judge. What he does recognise is that the decision not to include
estimates, because the Department felt that they were insufficiently
robust, was taken in good faith at the time. His view was that
it was a misjudgement.
106. So you do not think there was any failure
attached to the Alcohol and Tobacco Fraud Review, because Roques
does. You are rejecting that part of his report.
(Mr Broadbent) He believes that if we had included
estimates in the ATFR at the time it would have stimulated debate
and led to earlier action being taken. That is possibly correct.
The difficulty was that at the timeand he recognised this
was a judgement made in good faith at the timewe did not
actually have estimates. What was available at the time was an
estimate for a partial year, based on cases actually happening.
Not, as he says later in his report, a properly analysed, properly
evidence based estimate to which he attaches a lot of importance
in section 5 of his report on estimating.
107. The report is fairly damning about Customs
and Excise's ability to forecast. I quote here from pages 12 and
13, paragraphs 5.1 to 5.4. "A major factor reducing management
effectiveness in the collection of excise duties was the failure
to recognise and understand the reasons for the leakage which
arises in each tax regime. The increase in leakage of tobacco
revenues from less than £1 billion in 1996 to approximately
£2.5 billion in 1999", fairly quick and fairly large,
"was largely unmeasured and therefore undetected". He
goes on and essentially his verdict is that the ability to forecast
expected tax revenues was inadequate and ineffective. Why is that
so? Why is that? Is it because essentially if you are forecasting
leakage you are forecasting the failure of Customs and Excise
and that becomes a disincentive of the organisation to focus itself
on that process?
(Mr Broadbent) At the time the revenue estimating
was based on estimating the revenues the Chancellor could expect
the following year. That was the task the Department was set and
it discharged, and defined as that task the estimates are not
bad. What Mr Roques points out is that there was another task,
which it was not asked to do but which could have been done, which
was to estimate total theoretical revenue and then work out what
percentage we were collecting and that would lead to a different
set of tasks possibly being laid in the Department. We have made
it clear in our response that that is the better way to go and
that is increasingly where we are going.
108. Could you just explain which way you are
going now and what it is you are changing?
(Mr Broadbent) We are moving, and we have seen this
with the tobacco strategy, towards an approach where we shall
be estimating total potential revenue and in the PBRs we will
be publishing estimates for other taxes. For the first time we
shall be able to see the total theoretical tax take for the major
indirect taxes with the caveat I made about VAT. Already with
tobacco we have a target which is not to collect so much tax but
to get to a certain market share of tobacco. We could probably
anticipate over a period of timeit cannot be done overnightthat
other targets in relation to alcohol for example will be developed
in a similar form. Personally I welcome that approach.
109. What discussions have you had with Treasury
in relation to this difficulty of forecasting revenue in the areas
which have been prevalent? Have there been discussions with Treasury
about these matters?
(Mr Broadbent) There have been some discussions with
the Treasury, but it is less a case that the forecasts had errors
in them. The point made by Mr Roques is that the forecasts as
made were made to deliver the output requested, which was an estimate
for revenue next year. Over time, and particularly as fraud itself
has developed as an economic activity, both the Treasury and ourselves
have come to recognise that we need to spend more time estimating
total theoretical take and then how much we are collecting and
that is a development over the last few years which is going to
come to fruition in a couple of weeks' time.
110. From where do you get your information
on leakage? For instance, in my constituency for several years
in the late 1990s landlords were complaining of the unfair competition
from other landlords who were buying their spirits from men with
white vans at the back door and selling it at half the price it
was normally available at. It really was not a terribly covert
operation in Kent and parts of South East London.
(Mr Broadbent) Estimating fraud is very complex. I
distinguish estimating fraud from understanding and working out
what frauds are being perpetrated. Estimating fraud now starts
effectively from expenditure data. You need to find out how much
people are spending on, let us say, alcohol. That means going
into expenditure surveys, national statistical data and then working
back from that, deducting what is being exported, what is duty
free, what is shopped and you finish up with a residual which
with some work you can assume is a reasonable estimate of fraud
and that is broadly speaking the methodology you will see in the
paper we are going to publish. Distinguishing what fraud is being
perpetrated is actually both as difficult and quite different
and oddly enough I rather agree with you. We had to be guided
somewhat there by what is happening on the ground up. This is
one of the challenges. To work out what fraud is being perpetuated
on the organisation, does mean starting with operational data
very often and it is surprisingly challenging. This is something
we are now more aware of and have organised to make much more
rapid. The critical importance for an organisation like ours of
having big open communication cables from the bottom to the top
so that, although I would not guarantee it, I would hope that
if somebody who was working in our organisation spotted a fraud
he would come up the organisation rather quickly now. I shall
not bore you with the detail but it is quite an organisational
111. Are you just relying on statistics or do
you have investigators going out and spotting the white vans at
the back of pubs and that kind of thing, or spotting that Andorra
is getting so many cigarettes that the population would have died
of lung cancer two years ago if they had smoked them all?
(Mr Broadbent) In terms of understanding the nature
of the frauds being perpetrated, we do now a great deal of intelligence-based
work. In fact one of the changes we made in our reorganisation
was to bring together all our intelligence activities. That, together
with our investigative activities, does a range of work both here
and overseas which ranges for example from detailed discussions
with tobacco manufacturers, in the case of cigarettes, through
to test-shopping illicit goods.
112. If you have all that in place, where is
the difficulty in forecasting leakage?
(Mr Broadbent) It is being put in place now.
113. It was not in place.
(Mr Broadbent) It was not sufficiently focused. What
this does is allow us to understand more clearly the nature of
the fraud. The question of how to estimate leakage goes back to
an analytical exercise with consumption data. There are endless
issues with delays and lags but we have done enough work on that
to start publishing data in the PBR this year.
114. May I refer to one which is pending, which
you were warned about? Page 14, paragraph 4.8 of the Roques report
says, "There is some evidence within the NIS that fraud and
smuggling to avoid hydrocarbon oil taxes is on the increase. This
is particularly true in Northern Ireland but the problem is also
thought to exist in Great Britain. There is also evidence of growth
in fraud in the VAT regime. In neither case is early warning or
measurement of these problems emerging from analytical activities
and I suspect that they may become serious issues requiring urgent
attention long before the capacity to estimate the leakage has
been created". He is concerned about this. What are you doing
(Mr Broadbent) Two things. One is that I hope we have
done enough work since the Roques report that if Mr Roques does
us the honour of coming back to visit us again he might feel that
some of his concerns have been assuaged. I should make this point,
because I feel perhaps it did not come across in the debate we
were having earlier. I personally take this report very seriously.
I have been significantly guided by this. I remember saying to
Mr Roques right at the outset, "You hire the professional
help you need and do a job to give me a real insight into this
organisation, because I want that". It is valuable to me
and I have used a lot of this to make changes. I would never want
to go so far as to say that it will never happen again, things
are completely different. But I do feel directionally quite a
lot of change is under way. May I try to elaborate on this without
trying to pre-empt the PBR? The issue with oils is quite serious
and there are some quite serious issues with VAT. I believe the
fact that we know enough to be publishing later this month some
quite detailed numbers, both in relation to oils and in relation
to the area of VAT which I believe he is referring to, suggests
we have come quite a long way since he wrote those words.
115. What do you say about the co-operation
of the staff with Mr Roques who obviously appeared somewhat defensive
in that the chief investigating officer could not find a date
to meet him over the whole five months of his inquiry?
(Mr Broadbent) Mr McGregor was the one individual
who did not meet him and I said to Mr Roques that if he felt he
needed to see him, then I would not press him to meet his deadline
for the report if it was important. Mr Roques took the decision
in the end that it was not material. In terms of the co-operation
of the staff, everyone without exception whom he asked to meet,
met him and I am not aware there were any difficulties in those
exchanges. It is quite a frightening thing for civil servants
to go through this. They do feel they live in a very blameworthy
environment and they do not always feel they get a good hearing.
A certain amount of apprehension on their part is understandable.
116. You are saying, in relation to this hydrocarbon
and VAT question which is raised by Mr Roques, as I quoted, that
the organisation is now no longer resistant to finding out and
looking towards estimating leakage and that this problem which
has existed with tobacco and drink will not be repeated for hydrocarbons.
(Mr Broadbent) In terms of estimating they live and
breathe it now. In terms of repetition of the problem, I am more
wary. I have a lot of respect for the effort and the human and
financial capital which gets invested in exploiting fraudulent
opportunities. This problem will never go away. If we bring the
organisation to be more flexible and responsive, we can be a worthy
117. You referred to blame, but I think you
would accept that if £668 million is lost in four years somebody
should be blamed.
(Mr Broadbent) Yes, I think that we should take our
responsibility for that.
118. It appears that £340 million of that,
nearly half of that, was lost at one warehouse in the City. How
did that happen?
(Mr Broadbent) It is a large amount of money and it
may have been too large an amount of money in terms of some of
the issues which arise in individual investigations. Without wanting
to sound . . . the context is important. This is a very large
warehouse. We are talking about a warehouse which covers about
one million square feet with about 13 different units, with 3,000
customers, 8.5 million cases of spirits going through it.
119. £340 million is quite a large amount
of money to go walking.
(Mr Broadbent) The £340 million is over seven
years. You are probably talking on average over those seven years
of about ten per cent of the throughput, which is too high. I
doubt it was actually anything like ten per cent in more than
two years; it was probably a smaller proportion. The pattern of
this fraud peaked very sharply in two years. There were probably
two years when it was high and the rest of the time it was not
anything like ten per cent. With the volume of throughput, the
sheer size of this warehouse which was serving London, the number
of movements and the time span, which have to be taken into account
in making the judgements, it is not inconceivable this could have
happened. John Roques did look at a number of our investigations.
He had complete access to all our investigations, all the papers,
which I have to say was quite a sensitive thing to do. He examined
in detail a number of them and he does not make material criticisms
of the individual investigations.