Examination of witnesses (Questions 1
- 19)
TUESDAY 21 JULY 1998
RT HON
FRANK FIELD,
MP and MS
HILARY REYNOLDS
Chairman
1. Good afternoon, ladies and gentlemen,
could I declare the public evidence session on child benefit fraud
duly open. I welcome Howard Flight who, since 12 hours ago, has
been invited to join us and we are delighted to have him as an
able alternative to Nick Gibb, who has now moved on to the Treasury
Select Committee. Minister, welcome. You very kindly offered to
come and spend some time with us and I know your diary is full
and we are constrained to maybe an hour, an hour and a half. It
is very important work; the whole child benefit area in relation
to fraud is an essential part of our interest as well as the Government's
interest, and we are pleased now we have got the final report
of the Security Review that we can discuss that. You have helpfully
brought with you Hilary Reynolds, who is Head of the Security
Branch of the Benefits Agency, and we welcome her back from her
previous appearances in earlier versions of the Committee. I wondered,
Frank, if you wanted to say something at the beginning? The thing
we were getting worried aboutand some of these fears have
been assuaged but not all with the publication of the reviewwas
that it was taking so long. A recent parliamentary answer which
you helpfully put into Hansard suggested this has taken more than
twice as long as the other reviews. All other things being equal,
you would think that a child benefit fraud scrutiny would be actually
easier than some of the others. Why has it taken so long?
(Mr Field) Chairman, before I answer that can
I make a few introductory remarks?
2. Please do.
(Mr Field) I was anxious to come today because
it would be the first time that Hilary appeared before the Committee
when I was not sitting on the other side, and I thought it would
be rather nice to come and sit on the same side as Hilary for
a change. There are two main areas where you may wish to probe
methere are many areaswhere I would have difficulty
in answering. One is where there is commercial information relevant
to it and where contracts are involved, and though I wish to be
as open as I can, there are difficulties there. Secondly, while
many taxpayers may be watching this programme there will also
be fraudsters. On a previous programme under the previous administration
where counter-fraud staff appeared, when staff successfully busted
a ring the record of that programme was part of the training facilities
for that gang. So in some areas, if I may, I may not be able to
answer fully but I will supply you with the information. I am
anxious we keep some people guessing what we are doing rather
than telling them what we are doing. The final thing is about
how we might divide questions between us, Archy. At the end of
a very happy marriage Beatrice Webb was asked, "Why was your
marriage so happy?" She said, "Sidney takes all the
long-term decisions, I take all the unimportant decisions, and
I determine which are the long-term decisions." So if I may
be allowed to choose which are the difficult questions, and of
course take those myself, and give Hilary the easy ones, if that
would be all right, Chairman, we will go along those lines! You
asked a question about the length of time taken to produce the
report and there are two reasons why it has taken the length of
time it has. This report stems from a request that the previous
Committee asked of the previous Government. There are two reasons
why there has been a delay. Quite properly, when John Denham worked
on fraud he wanted to look at how fraud was defined and whether
in the keenness of an administration to chalk up fraud gains one
was putting in that category errors and other administrative difficulties
which ought not to be defined as fraud. As a result of the work
that John asked to be commissioned, not only within the Department
but with outside bodies as well including academics, we have a
much more robust view about fraud. Then, more recently, looking
at the report, I wanted to satisfy myself that the way we had
gone about the review was adequate and when I was satisfied it
was signed off, and we gave you the report as soon as possible.
It is late. It is not because there were difficulties, or unexpected
difficulties, it is because the new administration has taken a
much tougher line on how we define fraud, and had we had that
from the beginning the work you have been doing on the benefit
integrity project may not have been necessary. So it is all credit
to John Denham for actually initiating those big changes.
3. Are you reasonably satisfied with the
results? Do you think these results are accurate?
(Mr Field) They are certainlyas those of
the Committee who are statisticians will knowwithin the
bounds of reason in which we can read these results. What is crucially
important, Chairman, is what the main findings of the report are
and therefore where the Department now needs to concentrate more
attention, given that this was thought, I think by most of us,
as being the most secure benefit which we deliver and yet there
are some worries over the findings. Secondly, the Government obviously
wishes to accomplish much on this front but this is not the only
benefit it is looking at. Our counter-fraud proposals are themselves
part of the sort of first class public service we wish to deliver
for claimants, therefore we are as concerned about getting benefit
to people who are eligible and not claiming; as we are on cutting
down on fraud. So there are real priorities the Department has
to face in determining at any one time where it deploys its expertise.
Of the resources going to fraud, we have to think about how best
we can secure the taxpayer's advantage and within that we have
to decide what we are doing on child benefit.
4. Is there anything which will flow from
the report you have just produced? Is there any further work planned?
(Mr Field) There are two areas of concern and
maybe, as this is one of the difficult questions, Hilary would
like to come in as well! There are the 16-plus group and the findings
that the report shows up there. Secondly, the difficulties the
Department has had in the past in tracking people who rightly
move in and out of this country and do not always tell us when
they have gone. These are two areas which much concern me and
maybe Hilary would like to come in so it is not a monologue from
me.
(Ms Reynolds) The main areas of loss that we could
confirm through the survey were in the 16 to 19 year old age range,
children who were being claimed for as if they were still in full-time
education and were not, either because they had gone into training
schemes, gone and got work for themselves, in some cases signed
on for income support themselves and in some cases got married,
all of which render you ineligible for receiving child benefit
as well. So that was one key area for us. The second area of confirmed
losses was children in moving partnerships, where partners and
families are moving around and the care of the child also shifts,
and we have not always had the correct information there. There
is a wide range of other frauds which are set out in detail at
the back of the Benefit Review Report, which give you by numbers
how many confirmed frauds we have had in each possible category.
It is towards the end of the report, Table 21.
5. I would like to come back in detail on
that because colleagues have some questions on both of these areas
which you have identified as areas of continuing concern. While
Hilary has the floor, I wonder if you could just explain for us
in a wee bit more detail what the differences are between the
definitions of high suspicion of fraud, medium suspicion and low
suspicion of fraud, and maybe also a wee bit about how these savings
are calculated because some of the headline figures are taken
over a six month period in terms of quantifying what fraud is
extant? Could you help us with some technical definitions?
(Ms Reynolds) There are three categories of fraud
which we label in this review. The first is confirmed fraud, where
we have examined one of the random sample cases and found, through
either admission by the person or through third party verification,
that the basic entitlement conditions of child benefit have not
been met or are not now being met even though they were at the
start of the claim, where the customer could reasonably be expected
to have known they should have told us of the change and have
deliberately not done so, and where we have sufficient evidence
to support a prosecution, and that is quite important. So this
is a fraud where we believe we have enough evidence to prosecute
were we to go down that road. That is confirmed fraud. In those
cases we both stop the benefit and can recover a past incorrectly
paid benefit. High suspicion of fraud is where we as investigators
and we as adjudicators have a very strong suspicion that there
is fraud in there based on the evidence we have gathered through
the process. For instance, a case where somebody on a visit voluntarily
relinquishes their entitlement to a claim when we go forward would
fall into that high suspicion case, but the evidence we have gathered
would not be sufficient on its own to lead us to prosecution.
It is frequently around the family arrangements where we are told
that the partner only moved in that morning by coincidence just
as we were visiting. So there is a high suspicion there. Low suspicion
is based on the information we have gathered there could well
be fraud in there, such as the child might in fact not be living
with the person claiming the child benefit based on the type of
information we are getting back, say, from a questionnaire, or
we are unable to obtain third party verification from a GP or
a doctor because they have not responded to us either because
they choose not to respond or they feel uncomfortable in responding,
but there is not enough to go on to allow a prosecution, and it
is not a high suspicion. So there is judgment in there but there
is also a lot of expertise from our trained validators who have
looked at every single case.
6. That is distinct from incorrectness where
the report quite clearly shows there are some actual administrative
errors?
(Ms Reynolds) Yes, there is customer error and
there is departmental error.
7. And they are teased out and categorised
separately?
(Ms Reynolds) They are indeed, and that is a separate
and quite important issue for us.
8. The totality of the fraud estimated is
multiplied out by a half year, am I right about that?
(Ms Reynolds) For benefit reviews we take a snapshot,
so if the review says, "Let's look at all the sampled cases
and find out what has been incorrectly paid", we then annualise
that snapshot. So we are saying there are 234 cases of confirmed
fraud at around £10 a week, annualise that up and we get
the total annual loss from this benefit. It is not a weekly benefit
saving multiplier.
(Mr Field) Which is the number of weeks one. So
there is a difference.
Mr Wicks
9. Does that include those who have gone
home, as it were, back to their motherland? Because this is based
on a sample, is it not?
(Ms Reynolds) Yes.
10. Obviously that sample will not cover
those cases because they are abroad.
(Ms Reynolds) What it has covered is the total
sample we looked at, which is around 4,000, and the cases where
we have found we have lost money, and then that has been extrapolated
across the whole child benefit caseload, which is around 7 million
claims in a year, and annualised. So it will include, I think,
the four cases where the person has gone abroad. Those will be
incorporated in and we have lost as a result an amount of money
on those cases.
11. Yes, but that is from that random sample.
(Ms Reynolds) Yes, from the random sample, because
it is giving us a national overview without targeted
12. I am confused.
(Mr Field) That is a separate operation, Malcolm.
All these operations have these peculiar names but this one is
called Operation Rattle and is actually specifically looking at
those who have come to this country, legitimately claimed child
benefit and then have left this country and have not told us they
have left.
13. I understand that.
(Mr Field) If you look at those, then the Department
has actually dug up some very real difficulties there and we report
that. For example, you may wish to cite the evidence we did in
university towns where people have quite legitimately come over
here to study, claimed child benefit for their children and gone
back but have not actually mentioned the fact they have gone back
and been claiming benefit. Localised studies in those areas show
this is one area where we do need to be very careful with taxpayers'
money.
14. We have different data sources, there
is the random sample, there is the stuff about people going home,
as it were, abroad, but what I am searching for is what is your
best estimate of total child benefit fraud per annum.
(Ms Reynolds) Per annum the national level shows
£51 million lost every year.
(Mr Field) That is for those where we would prosecute.
If we then add in the cases where we are pretty certain, and where
practically in every instance people surrender their benefit books,
the figure goes up to £180 million in total.
Chairman
15. I would like to bring in Patricia Hewitt
in one second about the specifics of the report. I understand
that it is important to keep the methodology across these benefit
reviews consistent, that is absolutely well understood, but on
child benefit looking at the random cross sample it is a benefit
which has got £7 billion, £8 billion of expenditure
annually and what worries me is if you take the embarkation number
question, if people log a claim and go off, they can claim that
money for 16 years. So when you annualise it up, as you do for
other benefits, for the perfectly understandable methodology that
we all understand, it does not actually apply to that particular
scam, and you could, if you drilled down into the finer focused
abuse involved in that particular criminal act, be losing far,
far bigger sums of money than you report in your random sample
in the Benefits Review.
(Mr Field) We accept that. This is one area for
further work, both on the weekly benefits savings and the way
we compute costs for these benefit reviews. For example, if you
are claiming your mother's old age pension and she died 17 years
ago, it is not a weekly benefit saving of 30-odd weeks, and part
of the prosecution is trying to present carefully the information
on how long somebody has been defrauding the Department. We are
actually looking at this area following the Green Paper and it
may be something which the Committee wants also to consider further
and make representations on. There are the two areashow
do we compute weekly benefit savings for when we are across-the-board
properly trying to safeguard taxpayers' money by undertaking checks,
and we do that by weekly benefit savings, and then secondly, given
that we have an agreement with the Public Accounts Committee that
we are going to do these surveys on all our benefits one by one,
how should we in fact work out as Hilary was explaining the annual
loss. We maintain how that calculation is done so you could compare
one survey with another and, as you know, on income support we
have gone back a second time. That shows an increase but it may
be because we are better now at finding fraud than we were before,
so it is difficult to read too much into even the trend of the
figures. The only change we have made in how we present fraud
is that we are much clearer about the different categories. Hilary
explained those which we prosecute, where the person almost certainly
loses benefits, and the area where we have a suspicion of fraud
but it is no more than that. There are those three areas and then
there are the departmental errors and customer errors. We are
trying to present information to you now, to the House of Commons,
to the public, which makes that distinction between these categories.
It is quite important for us to make it clear, because we can
then see where the main areas of risk are for when we are working
out where we should put staff.
16. My instinct is that in these particular
very finely focused areas there is a much bigger potential for
fraud, putting it no higher than that, than is thrown up in the
survey.
(Ms Reynolds) Indeed. If you think of the sample
as your now national, annual snapshot across the whole customer
base of 7 million familiesthree-quarters of whom are not
on any other benefit and do not think of themselves as DSS recipientsthen
what we have is a snapshot of that benefit across the range. You
are right, we can then go into the risk categories, for instance
the 16 to 19 year olds is showing up as around a 10 per cent risk,
so this is giving us the big wide picture and what we can use
it for is to drill down to get a narrower picture with more risks
for us.
(Mr Field) One further point on that, the reason
why we are more confident about the 16 to 19 year olds is because
the previous Committee asked the last Government to undertake
a special review. First of all, the review itself was brought
forward and they particularly looked at 16 to 19 year olds because
that was the concern of the Committee. The results, both for that
special survey and what comes from this Benefits Review itself,
are very similar. I think what Malcolm was saying is that while
it is quite proper to read statistical significance into our results
for the whole of the sampleHilary said it was four families,
I thought it was two with eight children but anywayfor
those who were abroad, whatever it is, it is actually very small;
a cluster of families and you cannot read statistical significance
from that. What is worrying is the other work we have done from
Operation Rattle which shows, as Hilary was saying, if you dig
deeper down on some areas there ought to be major concern. There
is concern from the Government and maybe later on we will explain
some of the things we are doing, but not too much of them.
Ms Hewitt
17. Just sticking to fraud for a minute,
a high proportion of the fraud and the high and low suspicions
of fraud seem to be accounted for by living together or undeclared
marriages. Those, if I understand the tables correctly, entirely
relate to the lone parent premium, and given the change in the
structure of child benefit this will be a declining problem?
(Ms Reynolds) Certainly the problem we have found
on fraud within the lone parent premium is going to disappear,
yes, but we will still haveI am not sure it is a problemthe
trend that there are partnerships which change and children live
with different parents over time and you have the shifting around
of the care of the child and therefore the eligibility to claim
benefit on behalf of the child.
(Mr Field) The child may have two names, Patricia,
and therefore there might be two claims on them. Also, the way
we have built up the sample was to do part of the child benefit
survey as part of our income support survey, so there were issues
thrown up with not just so much one parent family benefit but
where there were other parts of income support claims which were
not right. Then we looked separately at those which were not on
income support and dealt with those separately.
18. Can you say a bit more about the difference
in the findings in relation to child benefit fraud but as between
parents who were also claiming income support and those who were
not?
(Ms Reynolds) It might be easier, to avoid getting
it wrong, if I could give you a written note with more detail
on that.[2]
Chairman
19. That would be helpful.
(Ms Reynolds) I would hate to get it wrong.
2 See Ev. pp. 28-29. Back
|