Examination of Witnesses (Questions 40
- 59)
WEDNESDAY 21 APRIL 1999
MR GEORGE
MCCORKELL,
MR PETE
SHARKEY and MR
BRIAN BARNES
40. Thank you for that. In that partnership,
and I think the partnership approach is right, is not one of the
problems the inevitable imbalance between the partners, in terms
of the expertise? The Chair, at the beginning, quite rightly,
described your own expertise, that is those sitting in front of
us today, as state of the art; is not part of the problem that
that state of the art expertise in these systems does not run
throughout the culture of a public body like the Department of
Social Security, and is not that the root cause of some of the
problems, that, inevitably, the private sector is going to be
at a different stage of understanding of some of these systems
than is perhaps the public sector, as a contractor?
(Mr McCorkell) I think, in terms of the expertise,
I was very pleased with the Chairman's remarks about our internal
expertise, and it is vital to the Department that we retain that
expertise; and you talked earlier about outsourcing, and I am
quite sure one of the things the Department is not going to outsource
is that expertise. We have a responsibility, and I have a responsibility,
in ITSA, to ensure that I maintain sufficient of that expertise
within the Department and within ITSA in order to allow us to
work with these suppliers, to make sure that we understand what
they are offering and to make sure that, in the end, we in the
Department retain control of what we get, that we get what we
want, as opposed to what the supplier might decide he wants to
give us. Now Peter may want to comment on this later, but that
process is quite vital to the future of ITSA and to the future
of the Department, so it is extremely important that we continue
to develop people with that expertise. It is perhaps, I hope,
not surprising to you that, although all three of us have worked
in Benefits Agency, all three of us have come through the mill
in IT as well; in years gone by, I was a plan programmer.
41. But how do you hang on to people, given
the goodies that the private sector can offer to people with that
expertise?
(Mr McCorkell) These are precisely the sorts of
issues that we have to tackle, within ITSA, (a) how do we create
that expertise, and (b) how do we hold on to that expertise. The
starting-point, and this, I believe, is good news, is that there
are people out there who like being public servants, even IT experts,
they believe that it is a valuable job to be a public servant
and they feel respected as a public servant, and even as an IT
expert they wish to stay as a public servant; so even though they
may well be offered significant enhancements to go and work elsewhere
they remain as public servants. I would like to think there are
three of them here, in front of you, who actually feel that way,
and we are public servants because of that. But that is not all
of it, and equally we have to look internally at our pay and awards
structures; we will never totally match the private sector, because
they can always add bidders, and it is not the nature of Civil
Service pay that you can do that. But we have specifically looked
at pay and rewards, and, again, Peter, perhaps you can remember
some of the detail on this, in order to change the structure of
the way we reward people, to make it more attractive to stay in-house
and to stay with us. I am not sure, Peter, if there is anything
else you would like to say on that.
(Mr Sharkey) Just a couple of points, if I may.
In which other job do you get such fun as this; it is a true reward.
On the IT expertise, or expertise in-house, just one point, in
addition. Yes, I think we have some expertise in ITSA, and we
are extremely proud, and it is both technological IT expertise
but also business expertise. Now, with working with the businesses
over the many years that many of us have, we actually know, I
would not say as much as they do, because that would be arrogant,
but an awful lot about how these businesses run and what their
drivers are, at a level that helps us a lot to identify what needs
they really have and to try to help re-engineer the business.
What the AFFINITY consortia and people like that bring in is hard-nose
IT expertise plus raw business expertise, and I think the most
powerful projects are when these things are merged together, so
you get a set of people who really understand the departmental
business and drivers, departmental IT current, because this is
not a greenfield site, you need to know what is going on. Also
within ITSA we have some modern IT expertise, we bring in powerful
consortia involving people like IBM, as well as the EDSs of the
world, you have got to have powerful technological injection,
plus a view of where the IT world generally is going; and I think
that is a very powerful combination. And when you mix that with
actual business people, as the MS1 project is, I think those are
the recipes for success; it is a bit like under PFI, what do we
bring to the party. On the pay and awards one, we had an issue
last year where we were hitting much higher attrition rates of
our staff, it is not a nice word, attrition rate, but they were
leaving in "droves" to go to the private sector. This
was caused by a few things. One, we had had fairly aggressive
pay settlements over the years, and people had found a kind of
pay drift between what we were paying and what the private market
was. The other one, and it has to be acknowledged, was the Year
2K thing coming along, where anybody who could debug COBOL programmes,
and we have lots of those, was extremely valuable around the world
to debug COBOL programmes and find everything with two characters;
so we lost quite a lot of people with that. The attrition rate
had hit 20 per cent of our people, which is unsustainable, you
are out of business, in time. We had discussions with various
people in the Department, etc., and we came up with what we thought
was a very imaginative pay settlement, I would not want to quote
the details here, if you are interested, it is not a secret, we
can send you the details, where we tried to target, in a performance-based
way, on a competency-based pay, rewarding people who had the skills
that we wanted to keep.[3]
This was argued and debated through both the Department and Treasury,
and I have to say that they were both very supportive, there was
a business case there, obviously, and some people got relatively
decent pay rises out of it, but it has not broken the generality
of the Treasury pay guidelines, it was just an imaginative settlement
that was supported by the Department, I was delighted with that,
and agreed to by Treasury. Since which, our attrition rate has
gone down to 5 per cent, which is about what we would expect,
through time. So, I think, on the pay and rewards, there is some
good news there, some imaginative work done last year which has
blocked that; there is still an attrition rate in IT generally,
I am sure you picked that up, the Year 2K is still a few months
away, as you know. So we are quite pleased with that, and all
the signals we get back from the staff are that that was a good
piece of work and they appreciate it and see that the Department,
ITSA, etc., made a stab at trying to make things better.
Chairman: I want to
move back to the core process with Kali, but, Doug, have you got
a quick supplementary?
Dr Naysmith
42. I have got a very quick one, that is
really only peripherally related to all this. But much of what
you have been saying, about managing projects, the modern way
of managing projects and involving private finance, sounds very
much, to me, like what has been said about smart procurement in
the Defence Procurement Agency. And I just wonder to what extent
you have been influenced by that, or maybe they have been influenced
by you, or whether there is something building up in the public
sector to look at this different way of doing projects, and if
there is any contact between you and, for instance, the Ministry
of Defence, or the Defence Procurement Agency, in the modern way
of handling projects and getting better results?
(Mr McCorkell) I will allow Brian to think, perhaps
he can remember. I personally am not aware of any particular contact,
but should there be, and I can check and let you know if there
was, it would not surprise me at all, because one of the things
that we like to say is that we wish to learn from anybody else,
if people have got lessons to teach us; the "not invented
here" syndrome does not apply to us. But these are difficult
issues and we must capture all the experience from everyone, so
it would not surprise me at all to find that we had talked to
Defence Procurement. I know we talk through CCTA to other Government
Departments. I know, for example, it was mentioned earlier, the
EDS relationship with ES and Inland Revenue, I personally will,
in fact I have already arranged meetings with my counterparts
in those organisations, so that we can share experience and so
that we can understand how we jointly manage a major contractor
who now works across all the Agencies. So I can assure you that
the general principle of (a) sharing our experience, but (b) us
using and capturing other people's experience, to help us develop
our strategies, is very well taken.
43. And there is no problem with the fact
that you have got two Agencies, which might have been considered
a little bit more competitive than collaborative in the recent
past, but it does not make any difference?
(Mr McCorkell) No. The whole focus now, across
Government, and this is not just in IT, is of Agencies and Departments
working together and jointly collaborating on projects, and, yes,
that is a relatively new and emerging scene for many of the Departments.
We have, I think, the advantage, in having gone through the JSA
experience, we had to work out a way of working with the Employment
Service, and that went through a long gestation period but came
to a very positive way of working. And we now all see the advantage
of sharing and working together, and we will be taking it forward.
(Mr Barnes) If I can just pick up the last point
as well. Working together is not a new thing for ITSA, in any
event, in that the legacy systems we have got were built with
a very high involvement of some of the leading IT people, so we
have worked in a different form of contractual arrangement with
them in the past. On the other point, about learning the lessons,
I think, from my own point, I was unfortunate in NIRS, in that
that was the very first PFI/IT, so I did try to see whether there
were any lessons to learn, and there were not. But, since then,
I have not had any direct contact with MoD, but a number of Departments
have visited me or I have been to them and given them talks and
presentations, so there is a general will there to share. And
I suspect it is more an indirect contact we get rather than a
direct, in that we feed our experience through CCTA and the Private
Finance Task Force, and they spread it out to other people, as
necessary.
Kali Mountford
44. Can I just start though by saying something
totally unrelated to procurement, and that is, it is good to hear
that some civil servants still think that public service matters,
and I was very pleased to hear your remarks about that, and, even
if we have to have a few inducements so that when we do develop
marketable skills people do not go off and market them, it was
very good to hear that the sense of public service still exists.
I notice, in your memorandum, there are actually 15 organisations
in total, if we add the three consortia together, involved, and
that obviously means that there is some sharing of risk, both
for them and for the Department; and, presumably, the thinking
behind that is that it lends the Department a great deal of flexibility,
ability to change systems and choose a relationship which is appropriate
to whatever project we are about to undertake. And I was listening
just to what you were saying about lessons learned; another lesson
surely to be learned is picked up in paragraph 3.6 here, and that
is the "Big Bang" approach and why we do not want it
any more. Have you got anything more that you can share with us
about the `big bang' approach, and why it is you have rejected
it, and what you see as the benefits of not choosing it now, and
what lessons have you learned perhaps from our past experiences?
(Mr McCorkell) Yes, we have very clearly rejected
the `big bang' approach, and I think that is for two reasons.
It is (a) the sheer practicality of implementing something as
huge as the total system; and, equally, in relation to the external
suppliers and the contracts, and Brian may again want to comment
on this, in contracting for something as huge as this over such
a long term and getting a set of contract conditions that you
then have to manage in a totally changing environment leads to
confrontation and leads to, you start with a contract that is
perfectly valid but in four years' time it is totally invalid
and you spend most of your time arguing about the contract rather
than delivering the business. In terms of delivering the business,
the key thing is to establish a piece of the business that is
consistent with your overall strategy and brings a clear benefit
and that is manageable in size to do, and Modern Service One is
our first tranche, our first phase, of that, and we will contract
for the development and delivery of that phase. We are now going
through the process of fully understanding that phase before we
write the contracts, and we are working with AFFINITY to get it
very clearly defined. We will then, and in fact we have already
started the process of saying, right, which is the next phase,
which is the next block of work that we will bring into this.
Modern Service One, I suspect, the decision on what that would
be was taken probably a year ago, was it, Peter?
(Mr Sharkey) It was finalised about a year ago.
(Mr McCorkell) Finalised about a year ago. So
we have had now another year before we even start thinking about
what the next phase is; so we can take the opportunity to look
at what has changed since we defined phase one, what is now the
priority, have the priorities changed, so we do not have to complete
the definition of the next phase too early, so we are better able
to intercept whatever the Department's real requirement is, as
the world changes and the world moves on. So it brings about those
advantages. It also brings about the advantages in contract terms,
that we know what we are contracting for, the supplier knows what
they are contracting for, and we can get a much firmer fix on
price and much more control on price. If you try to contract for
something huge then you start off with the contract and the price,
but as the world moves on and you change then you keep changing
the price and you have a tendency to lose control. Brian, is there
anything else you would like to say about the contracting strand?
(Mr Barnes) It is not directly contracting but
it is around the dangers of implementing systems with a legislative
deadline, where you almost come to a cliff edge, and if it is
not there on the day there is a major issue, and we have had a
number of our systems around that; so, certainly, we are looking
to avoid situations, wherever possible, that we get caught up
in very finite positions which give you no room for manoeuvre.
45. In the same paragraph, I think it is
even the same sentence, it also talks about pilots and trials.
Will the pilots and trials help us look at the end date and whether
we can meet it, or will it look at whether or not the system is
going to work; and I personally like having lots of pilots and
trials, I like testing things first, but, given that, are there
parts perhaps of the legislation on procurement which perhaps
give some constraints to pilots and trials, and, if so, how are
you going to deal with it?
(Mr McCorkell) I think, if I take first the pilots
and trials that we have already conducted, which were not so much
about how difficult is it to implement this technology, or how
long it will take, it was about is this the right type of technological
answer to the business requirement. So in things like Lewisham,
where we had a clear requirement to interface with local authorities,
then we could sit the IT people back in their black box somewhere
and design a brand new system for integrating Local Authorities
and Benefits Agency offices, but unless somebody actually goes
out there and tries to do it then you do not see the human interaction
and what things will work and what things will not work and the
best ways of doing it. So those trials were designed specifically
to inform the overall design of exactly what we will build, and
we will continue to use those sorts of trials and model office
type processes to try things out, so that we inform getting it
right, rather than we decide, we nice IT people, we are going
to have one of these and we go off and design it and build it
and then we take it down, and the benefit clerk says "Life
isn't like that." So they are very much designed to get the
design right and the specification right. In terms of can they
do it and can they do it in the timescale, the process that we
are now going through, working with them, is allowing them (a)
to be much clearer about what the requirement is, but they also
have to produce very detailed development plans, which, as opposed
to the way it happened on previous PFIs, we never saw; we now
see those, monitor those and agree with them that these are reasonable
and sensible, and hence we have confidence that it can be done.
46. I was going to ask you actually how
exactly the ACCORD differs from PFI, because I do not really fully
understand the difference. Can you explain that in a bit more
detail?
(Mr McCorkell) I think that, in terms of the change
in the nature of the relationship and the way that we will control
these projects, that is one of the big differences that give me
confidence that we will not allow the past problems to re-occur.
Because I believe it is true, in past PFIs, we gave the contractor
the specification, they told us they will do it by this date,
and off they went and started working, and what we saw then was
when they brought some stuff to test; that is not the way it will
work here. We will see their development plans, we will see their
progress reports against their development plans, we will be able
to monitor where they are, we have total insight into how the
whole thing is going. That is not saying we are managing it on
a day-to-day basis, but we have total insight into how it is operating
and where, potentially, problems may occur, that we can then sit
down with them and say "Hey, hold on a minute, unless something
is done about this we're going to end up being late", or
whatever.
47. Right; but in PFI one of the supposed
advantages, in any case, was that there was a transference of
risk to the private sector, and we heard from NIRS that perhaps
not all the risk went where we thought it would; but where does
the risk now lie?
(Mr McCorkell) The risk in the development, and
their input and cost of development, still lies with them, because
we will not pay until we get useful products; so the same risk
will still lie with the private sector partner. What we are doing,
through this process, is allowing us, in the Department, to manage
our risk, which we could not do because we did not know how they
were imposing on our risk. If one of these projects is delayed
then, in terms of the development, the private sector has taken
the risk because they do not get paid, and if it costs them more
money to do it they have to put that money in; so we have transferred
that risk, wonderful. But if it is delayed we have trouble telling
people what their pensions are, and not only, apart from the bad
public service, that costs us a lot of money, to either do it
clerically or put it right, or whatever. The difference is, we
are now getting sufficient insight into what they are doing, without
taking away their responsibility, to help us manage our risk and
make sure that things are going on track.
48. So is there going to be an advantage
to service delivery, or a disadvantage, indeed; how is this going
to affect the quality of service delivery for the Department?
(Mr McCorkell) In the way we are now working,
we are geared up to work in a way that will deliver the system
that we actually want and need, that we have tried out and piloted
in model offices, that the clerks out there know and can use and
can bring the benefits and will deliver to time. So, instead of
getting a system that is late and not quite what you wanted, and
all the problems that causes, we will get a system on time and
one that is what we want and what we are expecting, because we
are involved throughout the process. We still leave the risk of
investing in this development with the private sector.
49. And what about cost of delivery then,
the cost of the delivery of the system, perhaps we can see where
that lies, but are there any cost implications for service delivery
from the Department, I am thinking again of NIRS 2 perhaps?
(Mr McCorkell) In terms of the cost on the private
sector, I would argue very much that the cost of getting it right
and getting it up on time is actually much less than the cost
of getting it wrong and taking longer; so there is a major advantage
there. Clearly, there is an internal cost, in that we now have
to put more effort into doing this monitoring, to doing this input
that we keep needing to do on an ongoing basis; the cost of that
is a huge saving against, again, the cost of getting it wrong.
So, really, rather than put a lot of effort into correcting it,
it is put some effort into getting it right.
Ms Buck
50. I apologise, because I am going to ask
my questions and go. But, certainly, some of the questions I am
going to ask on the prototypes and pilots we have already covered,
and some of them have certainly lapped up against this, because
you have been giving very comprehensive replies; so feel free
to be as crisp you like, commensurate with accuracy. We talked
a little bit earlier about HORIZON and about the systems that
backed up HORIZON, the Personal Details Computer System; can you
just tell me, what difference does it make from the customers'
point of view?
(Mr McCorkell) Do you mean the HORIZON system?
51. Standardising the data system?
(Mr McCorkell) Standardising the data system,
let me give you an example. At the moment, if you are a pensioner
and you change your address, you might well write in to the Pensions
section and say "I've got a new address"; you might
then, a week later, need to claim Income Support, and go into
the Benefits Agency office, claiming Income Support, and they
will ask you "Do you live at 46 Arcadia Avenue?", and
you will say "Oh, no, I told you last week, I moved from
there." That will not happen any more, in terms of the direct
customer interface. And simple things, as part of the modernisation,
things that you will all have seen, if you go into Comet to buy
something, they ask you your post code, and that is all you have
to tell them, because they pull up the rest of the address; we
now do that, we did not before, because when our systems were
designed that piece of technology was not there, it has now been
implemented. There are some very simple things like that that
make it better.
52. Within the whole kind of process, there
is going to be built in, not just your job, but there will be
built in, I mean the risk attached to that system is that a single
error can then be duplicated all the way through the system; so
are you looking at how you can make sure that a name misspelling,
or something, does not—with a lot of ethnic minority names,
I frequently get a name misspelling and becoming a chain of errors
all the way through the system?
(Mr McCorkell) Yes, that was recognised in the
design of this system, and, as you say, that is mostly dealt with
with procedures in the front office and the people dealing directly
with the information as it comes in; and, as I said earlier, we
have trained 70,000 people to use this system. If it was just
a matter of "Well, here are some new screens and you use
a different screen", we would not have had to train them;
the main reason for having to train them was "this actually
makes your job different", because, although you are now
taking an address from a pensioner, it might actually be relating
to Income Support and you have to be more careful. There was a
lot of awareness training went on as well as procedural training.
53. Has the Data Protection Registrar been
involved, in terms of raising any issues about data?
(Mr McCorkell) We have checked with the Data Protection
Register, at the time that this system was designed, and it is
fully compliant with the requirements.
54. What has been the process of aligning
DSS data with Inland Revenue data, given the two or three very
large-scale projects or changes that would have affected the interface
between the DSS and the Inland Revenue?
(Mr McCorkell) I am not sure if either of my colleagues
would have any detail on that. Brian?
(Mr Barnes) I can give you some information, in
that we do interface various elements of data with the Inland
Revenue, and at the point of the interface there are certainly
agreed standard definitions to preserve those interfaces across
the boundaries. And, I think, in recent times, to say aligning
would be too strong a word, but certainly the process of sharing
the standards we each use to store names, address, marital status,
etc., is being shared, and that will be sort of taken into account
in any future exchanges between us.
55. Moving on. In your memorandum, you talked
about having had a very positive response to a number of the prototypes,
DSS Direct and the Camden Project, and the Lewisham Project; what
is that based on, that claim of being a very positive response,
and how are you using the monitoring to be constantly looking
at the design side?
(Mr McCorkell) Again, I will probably bring in
Peter here, because he had a very direct involvement in those.
But those series of prototypes, as well as informing the future
decisions, were reviewed quite carefully, and they did bring some
significant benefits; there were reductions in claims processing
times, there were reductions in the time it takes us to pass information
to the Local Authorities on Housing Benefit, there were identified
reductions in fraud and abuse, because we have speeded up the
process and passed more accurate information. There was a clear
response, I believe, from the public, in terms of our prototypes,
on what we call electronic claims processing, but, basically,
it is allowing somebody to ring in and we take the claim on the
`phone, rather than sending a huge form, and not only take the
claim on the `phone but if that is a claim to Income Support and
Housing Benefit it is done at the one time. There was a very positive
response to those.
56. And that was actually monitored——
(Mr McCorkell) And that was monitored and measured.
57. By what, by sending people response
forms to report on how they were perceiving a different quality
of service, or interviewing people?
(Mr McCorkell) I will ask Peter, who was involved.
(Mr Sharkey) A follow-up. I do not know exactly
who did it, but it was independent of us, anyway. The study was
commissioned actually by the Benefits Agency/the team who put
those prototypes in, and they are independent reports and assessments,
based on interviews, based on factual returns, based on, they
also had, I forget what you call it now, where you check the effect
on people who have not been affected, if you know what I mean.
So they were independent of us, and the results were, as George
says, mainly positive, in every aspect, both from the customers,
the clerks, or people working in either the Benefits Agency or
the local authorities, or the people just using the IT and saying
"Oh, that's good, isn't it, in colour"; that was a major
improvement for some.
58. And we are making sure, I know, again,
this question of the Benefits Agency as well as you, but we are
making sure that what could be an overwhelmingly positive response,
and I do not doubt that it is, and should be, to these schemes,
because they are very good, but an overwhelmingly positive response
across the board can still mask certain groups becoming actually
completely excluded, and that has just got to be borne in mind?
It does not undermine the validity of the good work, but it does
mean that you have got to look at, just because you get an 80
per cent response and the 20 per cent could be seriously marginalised,
as a consequence, you are really making sure that you are looking
at disabled people, you are looking at people whose first language
is not English, all those kinds of things, you are making sure
that whatever models you use across the board can encompass that?
(Mr McCorkell) I think the major protection there,
and, again, it is why we want to expand the use of these prototypes
and pilots, is the decision to do something in an entirely different
way and to stop doing something can only be taken when you have
done that full analysis and you know the total effect. So when
we say, for example, that, apart from informing the future, we
are looking at the possibility of implementing some of these ideas
and technologies elsewhere, that is not saying that we will immediately
replace everything else existing we do, we will take the advantage
of using these, but the other facilities will still be available.
The electronic claim form facility, if we make that available,
until we are absolutely certain nobody needs it, it will still
be possible to send in the written claim form.
59. The last question. The Lewisham and
Camden prototypes, you had to work with Local Authorities, in
one case, looking at their systems, Child Support Agency integration,
in the case of Camden; just briefly tell us what were the main
differences, what were the main problems you have overcome?
(Mr McCorkell) My understanding is that, in terms
of those prototypes, and perhaps it might be part of the fact
it was prototype, there was a very positive integration of the
people. The fear that the people in the Local Authorities, people
in Child Support, and in the Benefits Agency, would suddenly not
work together, because we would all be suspicious of one another,
just did not emerge at all and they all worked very positively
together, and there were no real issues. Again, I am not sure,
Peter, if you have any direct experience you might want to add
to that.
(Mr Sharkey) Just to confirm, almost uniformly
positive, both in the design stage and through the implementation
stage and the review stage; there were odd hiccups, obviously,
but in the main, from start to finish, complete co-operation.
(Mr McCorkell) I think, again, the thing to be
careful of, and it relates to your concern that we might miss
something, we do have to remember that these were prototypes,
and when you are doing prototypes people do tend to feel a bit
more important, because they have been selected to do something.
And, therefore, as we start to expand this into working, we must
be very careful about the communication to the staff and the understanding
of the staff, so that we do not raise undue concerns with them,
so that, basically, we help to create the same sort of atmosphere
and environment. And if we take the lessons that are coming out
of here, where the staff response was very positive, then I think
that is something to build on, to all of them, "This is going
to make your job much better."
3 See Ev. p. 27. Back
|
