Examination of Witness (Questions 1180-1199)
MR JOHNSTON
MCNEILL
15 JANUARY 2007
Q1180 Mr Drew: Take me through the
policy discussions. At what level were the policy discussions
taking place, who was involved in them and who made the decision?
Mr McNeill: The policy discussions,
as I mentioned earlier, involved Bill Duncan and Hugh Mackinnon
from the RPA. They were dealing with David Hunter from Defra who,
I think it is fair to say, was working to Andy Lebrecht, was taking
the lead in much of this work, as Andy was, of course, working
closely with ministers on it. I was not involved. I had no input
whatsoever bar providing the best expertise the Agency could,
officials who had been through the McSharry reforms, IACS, et
cetera.
Q1181 Mr Drew: This changes the whole
dynamic of the way in which the Single Payment System is going
to work. This goes for the most complicated model possible. You
have got to deliver this and yet you were not privy to those discussions
which actually said, "Look, Johnston, this is what we want
to do".
Mr McNeill: No, my expertise was
not in that area. I explained before that when I took up the position
I had not 30 years of CAP experience, I did not have CAP expertise
whatsoever, so I played to our best strengths and fielded the
best people we had. They were feeding back to us, the RPA Executive
Board, on a regular basis, we were taking sessions on our Executive
Board, they were feeding in initially that it looked like the
historic model was going to be it and that gave us great relief,
of course, but then it became clear that there was a move to this
much more complex system.
Q1182 Chairman: Could you just explain
when in timetable terms that move from historic to another model
started to evolve.
Mr McNeill: I am sorry, Chairman,
I do not have that information to hand. Perhaps by perusing the
Executive Board minutes I might be able to identify where we started
to grow concerned about the increased complexity.
Q1183 Mr Drew: But you did not have
any say at all on the decision of the model?
Mr McNeill: Personally?
Q1184 Mr Drew: Personally.
Mr McNeill: No.
Q1185 Mr Drew: No-one from the RPA?
Mr McNeill: I was never involved
in a discussion as to the policy development and where it was
heading, only to say at meetings, et cetera, that the more
complex the solution the more difficult it was going to be for
the RPA to implement and the more risk was involved.
Q1186 Mr Drew: So when Margaret Beckett
announced this decision on 22 April, how much prior knowledge
did you have of that decision?
Mr McNeill: We would have had
the knowledge from the feedback from Bill Duncan and Hugh Mackinnon
who would have been saying to us, "This is the line that
is going to ministers. This is the way it is looking and this
is likely to be what is going to be announced".
Q1187 Lynne Jones: There were people
in your Agency who had some input?
Mr McNeill: Absolutely. I have
mentioned three times Bill Duncan and Hugh Mackinnon who had been
involved in the previous CAP regimes. Hugh Mackinnon was our Director
of Operations, he dealt with 21 CAP schemes on a daily basis,
and had done for 30 years, and was now a senior civil servant.
Bill Duncan headed up our Central Scheme Management Unit, he was
our expert on all those schemes and had been involved, as I mentioned
earlier, in the McSharry reforms, IACS, et cetera.
Q1188 Lynne Jones: I have been trying
to find the record of it. We did have some information from the
NFU about a meeting at which both Defra and the RPA people were
present and one of them was attributed with the comments, in effect,
that it would be catastrophic to go down this particular route.
Mr McNeill: Yes.
Q1189 Lynne Jones: Yet that view
does not seem to have been conveyed to ministers. I cannot remember
who it was. Andy Lebrecht is Defra, is he not?
Mr McNeill: Yes.
Q1190 Lynne Jones: Does anybody here
remember who it was? It was Bill Duncan. Bill Duncan expressed
the view that it was a crazy choice to make.
Mr McNeill: Yes.
Q1191 Lynne Jones: Did he say that
to you? Did he say that as the representative of the RPA he was
telling them this was a daft choice to make but he was being overruled
by other people? Were you aware of any such discussion? Were you
aware that was his view?
Mr McNeill: Absolutely. We had
Bill Duncan and Hugh Mackinnon—Hugh Mackinnon was a standing
member of the RPA Executive Board—and we would have feedback
from Hugh at least on a weekly basis, if not through emails in
the interim, at EB on this issue, which was extremely important
to us: what is this SPS scheme going to look like and how complex
is it?
Q1192 Lynne Jones: So your guys were
telling the people that you thought this was a daft decision to
make in the first year at least.
Mr McNeill: At RPA EB we would
discuss the consequences. We would have Alan McDermott (RPA IT
Director) telling us about the IT consequences, we would have
Alex mentioning difficulties perhaps relating to finance and the
cost of developing that and we would have Simon Vry telling us
about the relationship with Accenture and what would happen there
and the increased complexity and the risk. Round the table[9]
we would have discussions and we would ask Bill and Hugh to continue
assisting Defra in this but to make it clear that we felt this
was increasingly high risk, which they did.
Q1193 Lynne Jones: Did you never seek,
as Chief Executive of that organisation, to perhaps talk to the
Permanent Secretary about these severe concerns within your Agency
from experienced staff that this would not be deliverable?
Mr McNeill: We at the RPA never
said it was not deliverable, we accept that. We said it was still
do-able. Nobody ever told us it was not do-able. Accenture in
their records do not say, "We told them it could not be done".
That has never been said. Karen Jordan never said, "This
cannot be done". OGC did not say, "This cannot be done".
It was do-able, it was just increasingly risky because of the
increased complexity.
Q1194 David Taylor: It was daft but
do-able.
Mr McNeill: It is not for me to
say. If ministers wanted to go down that road in the timescales
that they indicated, my job was to say if I, as SRO and RPA Accounting
Officer, had any information I could have put forward in an objective
way of some standing that said this was not do-able I most certainly
would have done so. I had a good working relationship with Sir
Brian Bender, who I respected greatly, and continue to do so,
and I would have been able to have had that discussion without
any difficulty.
Q1195 Lynne Jones: But high risk
implies it is a high risk of it not being do-able. If it is 70%
of risk, that is a 70% likelihood that it is not going to be do-able
and a 30% likelihood that it is going to be do-able.
Mr McNeill: Chairman, there has
been evidence in front of this Committee that we quoted probabilities
as low as 40% in terms of reporting on risk.
Q1196 Chairman: I come back to the
risk which the Agency attached to the processes which were being
considered by ministers. I would like to know whether the representatives
of the Agency, when you were comparing and contrasting the different
ways in which the Single Farm Payment System could have been introduced,
ie historic, hybrid or something else, and did ministers and/or
Lebrecht and/or Bender have a quantifiable risk rating according
to the choice of system. Looking at the timetable there must have
been a window of opportunity from when ministers agreed to adopt
the Single Payment Scheme under the reform, which was in June
2003, as opposed to 12 February 2004 when Margaret Beckett announced
the choice of the dynamic hybrid. There was roughly a six month
window of discussion opportunity. During that period did the RPA
quantify the various risks or the risk profile in a way that ministers
would have known it was X% for this and X% for that?
Mr McNeill: The team that was
fielded by the RPA was the best team we had on policy, and that
was Bill Duncan and Hugh Mackinnon. They were feeding back to
Defra the response from the RPA Executive Board, which involved
Simon Vry, Brian[10],
myself and others, and making it clear that it was increased risk
with increased complexity. Once the decisions had been taken,
such as the announcement that this was going to be a dynamic hybrid,
we sought as quickly as possible to understand the implications
of that and, indeed, as I recollect, produced a number of impact
assessments. I notice that Accenture say they were not consulted
in 2003 as to the consequences, I am sorry, I spoke with Simon
Vry on that and particularly remember sessions that we had with
Accenture. I think part of the difficulty here is that Accenture's
lead partner at that time, Barry Prince, is no longer with Accenture.
Following a review of the RPA change programme by Accenture Head
Office in America, a chap called David Hunter, I am afraid Barry
is no longer with Accenture and a new person has been put in.
Certainly my recollection is we did have those discussions, we
had to understand what the systems development issues were relating
to that, we would have been foolish not to. We started to look
at what would be the impact of the policy as we understood it
at the time but, Chairman, the policy continued to develop and,
in fact, continued to develop until the end of 2004 at which stage
we even had the OGC at the start of 2005 saying in their reports,
"Stop the policy changes". It was something of a moving
target as to what the impact or the risk was. It was difficult
to pin down the moving target because the SPS policy was changing
as we moved on until we reached the end of 2004. We did supply
impact assessments. We were giving the best estimates of risk
that we could at that time.
Chairman: I am mindful that James has
been waiting very patiently to come in and I would like him to
come in and then Peter.
James Duddridge: I will come in on the
issue of special advisers later.
Chairman: Okay, fine, as long as you
are happy.
Sir Peter Soulsby: You have been talking
about risk, and we have been talking about risk, but when you
were discussing risk were you talking about risk in terms of not
being able to deliver on time or were you discussing risk in terms
of what in fact happened, a catastrophic failure?
Q1197 Chairman: Can I add to that
because having been a minister the normal form, which I gather
has not changed significantly, is that the minister when they
have to make a decision gets a note and the risks are quantified
and there is a paragraph when it comes to the recommendation,
"Ministers should be aware that...." I am trying, and
I think Peter and my colleagues are trying, to understand upon
what basis was it still decided that it was within the realms
of achievable risks to go ahead with the most complex model where
policies at the time to go down the dynamic hybrid decision road
were not fully taken against a background of a Change Programme
that was evolving, against the background of a new piece of IT
architecture which had to be designed, in other words the number
of unknowns was increasing with the passage of time, and yet at
the beginning of the process the message I am getting is that
the decision to go down the dynamic hybrid route was taken as
a matter of principle without looking at this mounting tide of
risk and saying, "Have we got the risk/reward ratio right?"
Mr McNeill: To the best of my
recollection we were not asked to undertake sophisticated models
of what actually each part of this new scheme or each dimension
of this new scheme was going to be in terms of risk. The issue
which I believe was quoted through the evidence that you have
is the question was "is this do-able", and that would
seem to have been the issue. I do not recollect us constructing
any models. We fielded people to advise as to whether we felt
it was do-able. Can I just tell you where we were at that time,
Chairman? We were developing a new system for 21 schemes. In terms
of SPS there were nine separate schemes, bovine and arable schemes.
Some of those schemes were quite complex. Together that was a
massive piece of work. Initially, Chairman, we had been sold the
concept by Accenture that we were going to have a generic end-to-end
process system for a claim to pay with separate rules where we
could tailor the scheme almost without having to bundle miles
of code and rewrite and all the rest, but I am afraid it quickly
became apparent that with the packaged Oracle solution that we
had been sold by Accenture, which we were attracted to because
our financial system was Oracle and that would remove the risk
of the interface between the processing system and the financial
system which pays out the cheques, that would not turn out. So
we were in a position with Accenture where they had nine schemes
to develop and it was quite apparent that this was probably not
such a profitable contract, and you have discussions about were
they making money and all the rest. The fact was it became clear
at this price if they could not use that generic model this contract
was probably under-priced. With SPS along came the option of saying,
"We will not build those nine schemes, we will build one
scheme. It might be complex and it might be more difficult, but
when you add it up and look at it in terms of resources and in
terms of time and everything else, compared to nine separate schemes
to develop and the risk of getting any one of those wrong, it
is actually not unattractive". That is my recollection of
the discussions that we had with Accenture and others. They felt,
as I recollect, that this was do-able, that we already had a programme
office in place. That was another one of the reasons why we were
not violently opposed to the idea of going for SPS in 2005. We
had a world class supplier on the books, we had gone through an
OJEC procurement at considerable cost, we were in contract with
Accenture, CAP reform was covered in the contract because we had
recognised that whilst the CAP reform was going to come along
although, as Brian Bender noted, no-one expected it to be so fundamental
and such a massive change, but it was there, it was covered, so
we did not have to do another OJEC procurement, we had been through
the learning curve with Accenture, so when we looked at this,
even though it was complex and all the rest, we at the RPA thought,
"We have a team sitting here now ready to do this".
One of the big concerns, Chairman, in all these IT developments
is keeping the right people, keeping the right expertise, particularly
with your suppliers. If you let them walk off to another job you
do not see them again, they could be there for two or three years
on another programme. We had a team, we were in contract, we had
gone through the OJEC procurement, one scheme was considered,
even complex, to represent something we could achieve and that
was why we said it was do-able. What we were not clear about at
the time we were being asked those recommendations was that they
were going to take until November 2004 until we fully understood
what this scheme looked like with the added complexity that was
built on and built on down to things like fruit and veg, which
might sound simple, Chairman, but has proved to be inordinately
complex and difficult to reflect in the system's development.
Q1198 Lynne Jones: On that point,
one of the changes was when Margaret Beckett announced that there
would be three regions rather than two. You are complaining that
there were all these policy changes but did you have Bill Duncan
or anybody else in there actually saying, "Well, hold on
a minute, do you realise this is going to increase the risk"?
Did you attempt to keep things more stable and to stop those changes?
Mr McNeill: I do not think we
had the slightest doubt[11].
I have spoken to Bill Duncan, to Hugh Mackinnon and Simon Vry,
I do not think there was any doubt that we probably almost on
a daily basis were making the point, "You keep making this
more complex". We were looking at a programme here that had
red lights all over it.
Q1199 Lynne Jones: Who were you making
that point to?
Mr McNeill: The points were being
made at CAPRI where Accenture were asking when they would get
the actual information to enable them to get on.
9 Note by witness: "at the RPA Executive
Board". Back
10
Note by witness: delete "Brian" and insert "Alex
Kerr and Alan McDermott". Back
11
Note by witness: "that Defra knew of the increased
risk". Back
| 
