Examination of Witness (Questions 643-659)
MR MARK
ADDISON
28 JUNE 2006
Q643 Chairman: Good afternoon ladies
and gentlemen, welcome to a further evidence session of the Sub-Committee
of the Environment, Food and Rural Affairs Select Committee's
inquiry into the Rural Payments Agency. We welcome Mark Addison,
the former Acting Chief Executive of the Rural Payments Agency,
indeed a former, very senior member of the Defra Management Board.
We are very grateful to you for agreeing to come before the Committee
because I understand you are now an ex civil servant, is that
right?
Mr Addison: That is correct; I
am an ex civil servant from the end of May.
Q644 Chairman: We are particularly
grateful to you because one of the problems, as you may have gathered,
is that many of the people who were central to the decisions and
the processes involved in the Rural Payments Agency have moved
on for many and various reasons, so we are particularly grateful
to you for coming. I am almost inclined to start my line of questioning
with one very simple question. If you were able to answer it in
a detailed and comprehensive way, we could all go home early.
I shall put it this way to you. Can you tell us in your own words
what went wrong and why?
Mr Addison: I shall do my best.
For me there are two fundamental issues which explain why it went
wrong and then there is a variety of factors which I am not entirely
sure in my own mind were decisive, but they were probably relevant.
The two key factors which are responsible for the crisis and the
failure to make the payments to the published timetable were:
one, the volume of change and ambition that the RPA, with the
Department's agreement, accepted going way back to the original
change programme in 2001 and I shall explain that a bit more in
a moment. Linked to that were not simply the volume of things
they were trying to do, but the timetable to which they were trying
to do them and the fact that the deadline set by the European
Union was a completely inflexible deadline. Of course it is a
fixed point, delivery has to happen to that point and if you fail
to deliver at that point or in that way, you have a very big impact
on the customers that you are trying to serve. The combination
of the amount of things they were trying to do on the one hand,
the timetable on the other simply in the end was too much for
the organisation to be able to deliver against successfully. If
it would be helpful, I could mention one or two other factors
which were possibly contributory.
Q645 Chairman: You just keep talking
and you may be able to answer all the questions that we were ever
going to ask in one simple tour de force.
Mr Addison: We shall see. Four
or five other factors contributed to the difficulties. The first,
and this is linked to the timetable point, is that the fact the
timetable was very tight meant that the amount of testing—and
I know the Committee have covered this before—of whole-system
testing in particular was not as much as ideally it should have
been. Second, there was probably, I am not absolutely convinced
about this, but there was probably too fine a set of tolerances
built into the basic scheme design and the basic assurance mechanisms.
The training that was offered to frontline staff and to some extent
the managers, but particularly the frontline staff, could have
been stronger and better and deeper for a very new system that
was being introduced. The management information arrangements
and the focus of the performance data could have been improved
and could have been better fit for the purpose for which it was
being used. I believe that to be the case. Finally, as it turned
out, the task-based approach rather than the claim-based approach
was not so much flawed in original conception but, once it became
clear that the whole process was getting into difficulties, it
was a much more difficult set of processes to recover from and
manage than would have been the case if they had had a claim-based
approach from the first instance. I am not in a position to say
whether that judgment was right or wrong, but certainly once the
problems began to arise, it was more difficult to manage your
way out of it with a task-based rather than a claim-based approach.
Q646 Chairman: I presume that your
remarks reflect two things: one, the experience of this process
right back from the beginning of the change programme with the
RPA which goes back quite a long way—we commented on that
in October 2003; also the situation which you found when you assumed
responsibility after the removal of Johnston McNeill. I am also
aware, having studied the minutes of the Rural Payments Agency
Ownership Board, that you attended, certainly through 2005, all
of the key meetings and therefore would have been aware of these
things. One thing strikes me in what you have said. If, with the
benefit of hindsight, you have identified these key areas as your
major reasons for why things went wrong, does that not raise some
quite serious issues as to why some of these difficulties were
not spotted at the outset? It may have been complex, but Government
are used to doing complicated things. Yes, there were more applicants
under the new arrangements than the old, but MAFF/Defra had successfully
coped with the old IACS scheme, there was plenty of experience
around in Government, for dealing with complex IT things, not
always with the best results, but somehow, in spite of that pool
to draw on, you have identified some fairly fundamental failings
by management to spot the weaknesses in what was being proposed
by those who designed the system to deliver the single farm payment.
I just wondered what your observations were as to what went wrong
in that respect.
Mr Addison: You are absolutely
right about the basis of these views and I should like to be quite
careful and hedge them about with a certain amount of uncertainty
given the fact that I was with the RPA for two and a half months
and before that I was, as you rightly say, sitting on the RPA
Ownership Board and latterly the Executive Review Group, so I
had some view of what was going on, but not a detailed view. It
is good to have a chance to enter that caveat. I was trying to
distinguish, in my explanation of the reasons why things went
wrong, between these two fundamental points which I shall explain
a little bit and these other issues which were probably not individually
decisive and every project and every programme has a number of
things which could be done differently and better, which is why
one has all the gateway review processes and evaluations at the
end of them. They were not the decisive points. The decisive point
was this combination of overall ambition for change and the timetable
to which it was being done. On that, it was not actually so much
the numbers of claims, the volume of SPS activity in that sense.
I know the Committee have spent some time working through the
numbers of new applicants, whether they were expected or not,
how big an issue that was, the number of changes to parcels of
land and so on. We could come back to that if you like and those
are certainly worth some discussion. What I meant by the volume
of change activity was the combination of three key programmes
simultaneously, so it is good to have a chance to explain that.
The first was the original RPA change programme, which was the
one that was launched in 2001 that Accenture were brought on board
to help with. It involved the introduction of a new IT system.
It involved the switch from a regionally-based organisation to
a national organisation with sites which did not deal with local
regions but certain sets of tasks, the reduction in the number
of sites by three down to five and it was designed to deliver
some very substantial savings to a certain timetable. That was
the first big element of this three-fold programme to which I
am referring. The second was clearly the CAP reform and the changes
that needed to be made as a result of that to the system design
and the nature of the payments to farmers. Linked to the second
was the third, which is the rural land register which was launched
in September 2004 and being built, used, and deployed for the
first time in parallel with the single payment scheme. In my view,
it was that combination of the three programmes together which
in the end was an overwhelming set of requirements for the RPA
to meet and as a result it failed to do so. The decision that
was taken to run all those in parallel was of course taken some
way back. It was taken back in 2003 and Sir Brian Bender himself
referred to this in the evidence that he gave you relatively recently.
Q647 Chairman: The thing that I find
difficult, before we explore some of the detail to which you have
been kind enough to refer, was the heroic public representation
that somehow, in spite of all of the things that you have talked
about happening or not happening, as the case may be, the heroic
statements about when things were going to happen in the form
of the timetable for payments. At the end of the day that is what
the revised RPA was tasked to do and in spite of a growing awareness
I am sure—perhaps you can confirm that there was a growing
awareness that perhaps you had bitten off a bit too much—people
were still saying almost right up to the bitter end that everything
would be alright on the night. For example, when the Lord Bach
came before the Committee, he was very vehement in his defence
of the timetable and in fact was very critical of the Committee's
report which had the temerity to question whether it would actually
happen and whether in fact there should be arrangements for interim
payments. Perhaps you would like to comment then on what informs
this sort of heroic public persona, when in actual fact some of
the problems which you have adverted to clearly must have been
becoming more self-evident with the passing of every day during
2005?
Mr Addison: There are two rather
separate questions here. Taking a step back, the first question
you raised was why it did not work and that does go back to the
degree of ambition originally. How any project of that scale or
ambition is going to go wrong is impossible to predict at the
outset. One just has a sense that somehow or other that is biting
off more than can be successfully chewed. It is a rather different
question to say: why was there not a sharper understanding, even
quite late on, that it was not going to work? That is a more difficult
question to answer and I should say there is a number of factors
here. Let me start by saying that it is a legitimate question
to raise and indeed criticism of all of us that were involved
with the process that we did not identify some of these underlying
issues earlier than we did. The reasons for that, which I shall
try to explain, were as follows. First of all, remember what the
RPA at the time was trying to achieve: it had a massive number
of different elements of the programme to get right; it had the
customer register to sort out; it had so-called level one validation,
the basic checks, to sort out; it had the document management
processes to sort out; it had the cross-compliance inspection
functions to sort out; it had a very wide range of things to resolve
as well as what turned out to be the decisive indicator, which
was the rate of level two validation so-called. Forgive me if
I am getting a bit into the technicalities. Level two validation
is the main bit of the process after the basic checks have been
completed in level one. The claims go into what is called level
two validation and the important checks are successfully completed
to make sure that the claim is right and accurate and that the
information on land in the claim matches up with the information
on the rural land register. That was one component, so the RPA
had an awful lot of other things to get straight and therefore
it did not give, in the end, the rate of improvement of level
two validation as much attention as it probably should have had.
There were maybe two other reasons. Putting the best possible
interpretation on it, if one looks at the position in mid-March
of this year, 44,000 claims had successfully completed level two
validation. If one takes the bulk of the payments target by the
end of March, on the most generous possible assessment to the
RPA, to be 51%, just over half in other words, then they were
roughly 16,000 claims short of that in terms of level two validation,
so they were off. It must have been becoming clear, if the right
figures had been looked at, that they were off, but they were
not a million miles away. That was one factor. The second factor
was a consequence of the task-based approach and it is worth a
bit of explanation even though it sounds a bit technical. The
task-based approach involves breaking down each claim into a number
of separate tasks and completing all of those tasks and when all
of the tasks on a claim are complete, then the claim is cleared
and it successfully passes level two validation. The task-based
system therefore inherently produces a model, an expected profile
of delivery which is end-loaded. With a task-based approach you
do not expect to see a huge amount of progress early on because
you are doing a lot of tasks, not necessarily specifically with
the aim of getting individual claims successfully concluded, you
are just working on the tasks. So you might find that you have
40 claims, you have 10 tasks on each and you only maybe do one
or two tasks on each, so no claim is successfully processed to
completion in that period. As the balls begin to fall into the
holes, you expect the profile to start rising very sharply and
you expect most of the successful level two validations to come
at the end of the process, not the beginning. This is a feature
of the forecasting model that a task-based approach entails. It
is therefore quite a risky mechanism to use, because if nothing
much is happening, it is quite hard to tell whether nothing much
is really happening or whether everything is going as it should
and you expect to see a rapid amount of movement towards the end
of the timetable. Because that was the model and that was the
approach, there was a sense in which the RPA management team would
not have been as exercised as some others about relatively low
rates of progress on level two validation earlier on in the window.
Maybe that helps to explain a little bit why they remained more
confident than outside customers and outside commentators that
it would come alright on the night. I agree, if one does not have
some explanation of that kind, that it is quite hard to see why
the level of confidence internally which I, as a member of the
Defra Management Board, had previously shared in the autumn of
2005, should have been there. Nonetheless I suspect, had a more
detailed analysis been undertaken over a period of the level two
validation rate, it might have been clearer that, even allowing
for the way in which the model worked, it was ambitious to hope
for results by the end of March. Nonetheless that was the message
coming out the RPA and which the Department and ministers were
responding to.
Q648 Chairman: I suppose what you
seem to be saying is that, having opened up the box labelled RPA
and looked in, you have drawn these conclusions, but that the
flow of information throughout the whole of the process, from
the start of the change process through to the preparations for
the single farm payment, the feedback mechanism to the Ownership
Board and to Defra's Management Board and subsequently to ministers,
the gift of that information was very much in the hands of Johnston
McNeill and his senior colleagues and you, as the "outsider"
representatives, because you and Andy Lebrecht were key senior
Defra officials on that Ownership Board, had effectively to accept
at face value what you were being told. You did not have at your
disposal, either any independent or Defra technical advice to
say "You guys are not IT experts, you are not systems experts,
you are management, you are officials and I think there might
be something there that needs a bit more probing".
Mr Addison: I would not accept
that the Department was completely reliant on RPA advice with
no kind of independent purchase. As you have heard, the mechanisms
that were set up to govern this programme were extensive, they
had independent input and they had hefty independent input both
from the non-executive director on the RITA Programme Board plus
the OGC reviews and indeed some other work that was done to check
on progress. It is not fair to say that the Department simply
had to take what the RPA told it was going on. The RPA was as
keen as us to make sure we had some independent view about the
progress we were making. On the question of looking back and what
I would have done differently, with the benefit of hindsight—and
everything I have said is with the benefit of hindsight because,
as you rightly pointed out, I was part of the departmental governance
mechanism—two things and one of them bears very directly
on your point. The first is that at the decision point when the
Department was being advised by the RPA that these three elements
of the programme could be run together, it would have been helpful
to have had an independent view of somebody who was not actually
that closely engaged with the programme, but was familiar with
large-scale financial retail delivery or systems or something
of that sort, who could have come in and given us a sense of the
degree of risk that we were taking by trying to run those three
programmes together. That might have been helpful. The other thing
that might have been helpful—your point about having some
stronger capacity to interrogate and challenge the assessments
that were coming from the RPA—would have been if we had
all, that is the RPA and the Department, had a stronger purchase
on the whole system and how it worked. As it turned out the RPA
did not have a strong enough capability itself of that kind and
the Department therefore was not able to share it. That is one
of the things I know that is being looked at right now and which
we were quite keen to try to build subsequently: a stronger understanding
of the way in which the different components of this new work
process and IT system were going to relate together in order to
deliver results. That capability was not strong enough as it turned
out, either on the Department's side or the RPA's.
Q649 Chairman: When Margaret Beckett,
I presume, summoned you to take over and try to sort the ensuing
mess out after she had decided she had lost confidence in Mr McNeill,
and you went to see her, what did she tell you that you ought
to do or did she just say "Mark, go and fix it"?
Mr Addison: The remit that I was
given flowed very directly from the immediate cause of the crisis,
which was that the expected rate of flow of payments for 2005
claims had not materialised, that the end-March target was going
to be missed and that the Department's and ministers' fundamental
priority, not just for the RPA but for the whole Department, was
to complete the process of single payment scheme payments for
2005 as fast as was legally possible. That was the remit that
I was given and that was the remit that I took away.
Q650 Chairman: The thing that we
found quite intriguing was that in a relatively short space of
time, you came in, you announced a series of measures that amounted
to a fundamental restructuring of the operation and magically
money started to flow. You then brought in the interim payment
mechanism which, according to the minutes of the Ownership Board,
you had been considering for some considerable time but had dismissed
up until that moment. We were left with a sort of situation of
saying "How come this guy Addison goes in there with a magic
wand and makes it all happen when after the previous two years
of so-called meticulous planning, the end result was the dismissal
of the chief executive and ministers being left to hang out to
dry?". You came in and sorted things out within a few weeks.
It was something that did not really fit very well together. That
is the point: that you came in quite incisively and it would be
very interesting to have your commentary as to why you made the
changes that you did. If you were able to produce that analysis
in a relatively short space of time, how was it that those who
were responsible for the RPA's management for a longer period
of time could not see what you saw effectively in a matter of
days?
Mr Addison: The first thing to
be clear about is what was achieved during the period I was at
the RPA and what was not and I shall happily run through that.
Your summary of it, although obviously part of me would like to
say "Yes, it is absolutely right, I did achieve all those
wonderful things", I am afraid is an exaggeration of what
was actually achieved during my period there. Certainly some things
improved and some progress was made, but in other areas, the problems
remain and it would be good to explain that. Then there is the
question about why, if it was possible to do that, was the problem
not identified earlier and something done about it earlier? On
what was achieved and what was not, there were several phases
which I personally and the organisation went through after Johnston
left and I arrived in terms of analysing the problem and identifying
what needed to be done. It might just be useful to run through
about five. The first was the one that we were just talking about
which is why, when the button was pressed, only a few payments
seemed to emerge. What was the reason, in Helen's words, for the
gumming up of the process? This did genuinely take everybody by
surprise. There had been an expectation that once claims had successfully
passed level two validation, once the button was pressed to make
the payments, there should be very few hold-ups and they should
flow through. In other words, at the end of the pipe there should
be no blockages. It turned out that there were some quite significant
blockages at the very end of the pipe which prevented the 44,000
claims which had been successfully sorted being translated into
payments. That turned out to be about a number of what are called
authorisation checks, final pre-payment checks which are done
before the claim is passed over to the payment engine for payment.
In all of these decisions we were obviously very keen to involve
the people who ran the actual business and operation in the RPA,
the internal audit people and the lawyers to make sure what we
were doing was sound and the risks were properly identified. Over
that very first weekend the RPA experts recommended abandoning
four of the six so-called authorisation checks and subsequently
in fact we realised we could abandon five out of the six. We took
the view they were redundant, they were already covered by the
checking process as part of the mainstream level two validation
and we cut them out. That had a big effect: when that claim was
checked, it did not just stop that individual claim as by then
the claims had been batched up and the whole batch would be held
up, so it amplified the effect. By removing those checks therefore,
you did not just allow those individual claims to flow through
you allowed the batch to flow through. This is a slightly simplified
version, but that is basically what happened. That was the authorisation
check stage.
Q651 Mr Williams: Could you just
tell us what sort of checks those six were?
Mr Addison: Absolutely. I happen
to have a list. The first one, which was the one we retained,
is the payment being made to the correct payee, in other words
a final check to make sure that the person you are going to pay
the cheque to is the person on your records. Just some examples
of the remaining five. Check two: are there any obvious reasons
why the payment cannot be made? Check three: is the claim for
an eligible scheme activity? In other words, is the scheme activity
being claimed for eligible? Check six: the amount that had been
calculated appears reasonable and conforms to scheme rules. In
other words these are quite general checks; you are asking the
operator at the end of the process to confirm finally that everything
is in order. They are quite general and they are quite discretionary.
There was quite a long routine of steps that individuals had to
go through, but the view that the experts were able to take was
that they were already covered by the main processing machine.
Q652 Mr Williams: Were those checks
carried out by anybody or were they part of the computer system?
Mr Addison: No, they were carried
out by individuals. The computer system, the RITA system, allowed
them to be undertaken and allowed the result of them to be recorded
and then the necessary steps to hold up that batch of payments
would flow. They were conducted by individuals but they were done
on the system. That was the first step. It was very clear quite
quickly that that was not going to solve the fundamental problem.
That would resolve the immediate cause which precipitated the
crisis at the point where it became clear that the payments that
were already expected to be sorted were not sorted, therefore
the end-March timetable became impossible. The more fundamental
problem was the rate of progress on level two validation. How
quickly were claims getting into that bit of the pipe? That was
clearly moving too slowly and almost whatever metric or measurement
you looked at and tried to forecast by simply extrapolating from
what had been going on in recent weeks or months, it was pretty
clear that that would fail to hit the end-March deadline and probably
fail to hit, on most of the measurements, the end-June deadline.
So there was a fundamental issue with level two validation. The
next thing that became clear was, for the reasons that I have
described, that the nature of the model meant that forecasting
was immensely difficult and we also, of course, had had our fingers
very severely burned by the failure to get the forecast right
in the past so it seemed to me quite early on that forecasting
anything would be very high risk. We simply did not have the data
or evidence to forecast confidently pretty well anything and,
as a result, for the next few weeks, indeed even now, ministers
are taking a view that it is very difficult to forecast ahead
timetables for completing certain phases and that is absolutely
right, that was my experience. You mentioned partial payments.
This was the next issue to address. A system had been set up for
partial payments and it was decided not to deploy it in January
or February when it became clear that actually the payments should
start in February. The phasing on partial payments worked as follows.
For the first two weeks I and my RPA Executive Board colleagues
took the view that partial payments would be a distraction; we
could not afford to take people off the main processing to design
or to take a fresh look at partial payments. Once those first
two weeks had elapsed, it became pretty clear that we could not
offer ministers sufficient confidence that we could hit the end-June
deadline for completing the level two validation process and we
therefore began to plan in earnest. After four weeks I felt that
we had got to the point where we were clear that a suitable partial
payment solution—that is quite an important qualification:
a suitable partial payment system—could be deployed that
would work. Ministers on the back of that announced that they
would deploy it as soon as it was operationally possible to do
so and in six or seven weeks we had deployed it. That was the
sequence of events on partial payments. We started with no work
and ended up deciding to do it when the full difficulty that we
faced became clear. Finally, in terms of the steps that were then
taken to try to resolve the fundamental problems, we took the
decision to move from the task-based system to the claim-based
system. We opened up communication between processors and individual
customers by telephone and we decided to bring the mapping back
in-house in order to create a package of measures that we hoped
would speed up level two validation. The conclusion we reached
on level two validation was that there were no simple silver bullets.
There were no obvious steps we could take to remove elements of
it that would radically speed it up. It had to be a redesigned
work process and I believe those fundamental challenges and difficulties
remain. The system remains and will remain for some time, difficult
to run, complicated and in some respects quite clunky because
a number of manual and off-system work-arounds, so-called, were
put in place which affect productivity and make it harder to run
the whole arrangement. Those problems remain. We could not solve
those overnight, but we did think a new approach to doing the
work would help.
Q653 Mr Williams: When Accenture
were first contracted by RPA, it was to deliver the change programme
and to update the IACS systems. It was only when the 2003 reform
came into place that they were then tasked to produce RITA, the
system to deliver the single farm payment and also to do the rural
land register. We were told by Accenture that dealing with those
three separate tasks was not an exceptional request to be made
of an organisation like that. Are there any other examples in
the public sector of a contract being upped to such an extent
after the contract had been entered into or do you think it is
exceptional?
Mr Addison: I am not sure I can
give you a definitive answer because I would not know enough myself
about the range of examples there might be. My sense would be
that it was an unusual amount of change to introduce shortly after
a contract had been signed. It was a very fundamental change.
Mr Williams: We have recently been to
Germany and met with some officials there who delivered the single
farm payment. They did it on a regional basis but it was basically
a dynamic hybrid just as the English system is. Each of the La[lcodot]nder
was allowed to commission its own software and yet each managed
to deliver the system on time and within the window.
Q654 Chairman: There were 353,000
customers.
Mr Addison: On that, I am not
sure I know the answer but a relevant question would be: at what
stage of development was the digital rural land register when
the new software was introduced, alongside it presumably? I do
not know the answer to that but that was quite an important issue
for England. The other factor goes back to where I started, which
is that the policy issues were clearly one set of issues. I am
not sure whether the dynamic hybrid in itself... I know the Committee
have discussed whether delays of one year or whatever would have
made a difference and I am not sure of the answer to that because
I do not know what policy freedom there would have been. What
I am saying to you is that it was the combination of the RPA change
programme which, remember, was not, I assume, something which
would be replicated necessarily in other countries, it was a completely
new kind of organisation being deployed alongside a completely
new policy, alongside a completely new rural land register which
was the factor which made the whole thing topple over in the end.
I am not sure how Germany compares with that or any other country
compares with it. The dynamic hybrid component is one of those
three.
Q655 Mr Williams: The mapping issue
and the rural land register seem to be key factors in this. This
afternoon we were given an example of a particular customer of
the RPA who had 19 different maps delivered to them before the
correct solution was found and that was in May of this year. We
were told by the RPA that one of the difficulties was that every
single piece of land had to be mapped before the total land area
had been identified and the entitlements could be worked out and
yet payments started to be made before this particular customer's
land had been mapped. One of the other issues that we were talking
about was whether the RPA and the system that they were using
was too accurate and that a little bit less accuracy could be
accepted to get the system working. Do you have any comments on
that?
Mr Addison: I mentioned earlier
on that one factor which is worth exploring, I am not sure it
was a decisive factor, but it is worth exploring and thinking
about, is whether the approach was not too finely tuned. One of
the steps that the RPA took after I arrived was to introduce a
tolerance limit of two hectares or 3% on individual claims; if
that tolerance was not exceeded, but there was an issue and the
system was saying there was a problem, nonetheless the payment
was to be made. We did quite a lot of calculations on varying
that level of tolerance and seeing what the impact would be on
the number of claims that would go through and the potential level
of disallowance that would be entailed. Obviously that is a factor
with all these decisions and that was why we ended up with those
figures. Interestingly, although it helped, it made some improvement,
it did not make the enormous improvement and change that you would
have expected if the fundamental reason behind all the difficulties
had been the degree of tolerance. That rather supports my feeling
that that was a factor, but not a decisive factor in the whole
set of reasons why the target was not met.
Q656 Mr Williams: After meeting with
the German officials, it was quite clear that they had got the
money out to the farmers. The question really is whether they
got it out in compliance with the European regulation. We shall
only know that presumably when the system is audited and there
could, of course, be disallowances as a result of that there,
as there may be in the English system. Has Defra, as a result
of the less fine measurement of the land areas, made any allowance
for disallowance in its budget in the future? Was that built in
as a result of taking the decisions that you had to take?
Mr Addison: The position, after
it became clear that the target was not going to be met, meant
that the whole approach to calibrating risk had to change. Ministers
in the Department were extremely supportive throughout my period—I
am sure they are still being extremely supportive—and willing
to consider options which would have been difficult to consider
six months ago and those options do entail, because nearly every
change does, some additional risk of disallowance. We tried to
advise them as best we could on the range of possibilities and
the decisions were taken. The whole question of disallowance is
not, as you will know, an exact science. At every step we took
the best advice we could about how to limit its impact and we
were extremely careful with the partial payment system which was
not the original design, it was a new design in order to meet
the requirement of the time. We were extremely careful to make
sure we had as many controls as would enable us to argue and negotiate
the level of disallowance down as best we could. We had the arrangements
for claw-back: if a payment had been incorrectly made that was
more than the payment that was subsequently validated, then we
shall seek recovery and that is exactly what is happening. We
have had relatively few in percentage terms, but that system is
in place. You can introduce a number of steps to try to reduce
the risk, but you have to have some appetite for risk, otherwise
you cannot move.
Q657 Mr Williams: Presumably the
rural land register process was carried out in Wales as well even
though the payment was on an historic basis. We seem to have had
very few complaints in Wales that payments have been held up because
the land had not been mapped. I do not know what system was used
in Wales because there have been so few complaints. Could you
tell us whether a different system was used in Wales or a similar
system?
Mr Addison: The policy was different.
What I am not clear about, I am afraid, and what I cannot help
with, though I am sure we could get this information for you if
we asked the Department, is how far the land issues had to be
resolved in order for those payments to be made. I am sure we
could find that out.
Q658 Lynne Jones: As far as you are
aware there is no difference between the mapping system in Wales
and England.
Mr Addison: The rural land registry
itself? I am not sure, to be perfectly honest, how Wales have
set about mapping. I do not know whether they are using the rural
land register we use in England or not.
Q659 Sir Peter Soulsby: May I just
take you back a little bit to your description, with the benefit
of hindsight, of the things which could have been done better?
You have described them to us and it was clear from your description
of them that none of them was individually fundamental. You also
described what you termed as the decisive points, the volume of
change and the timetable and its inflexibility. They had been
factors that you had been aware of for quite a long time. Even
the volume of claims was something that had been known for a substantial
time. What I still find very hard to understand is how it was
possible for all of those involved, the Rural Payments Agency
and the Department and the ministers, to be so apparently confident
right up to mid March and to be telling us how confident they
were, that this could be delivered and then so suddenly for matters
to emerge and for the position to change so dramatically; suddenly
it was not going to happen. Was it that the chief executive and
the others in the Rural Payments Agency did not realise that this
was not going to deliver or was it that they realised but were
not telling you and us?
Mr Addison: There were these two
factors. The issue which took everybody by surprise, including
everybody in the Rural Payments Agency, was the failure of the
44,000 claims to go through into payment smoothly. The authorisation
checks' element of the process had not been tested as part of
the whole system test, a point raised earlier. That simply did
take the RPA by surprise.
|
