HMT00000025
HMT00000025
a
CHIEF SECRETARY
RESTRICTED - POLICY ‘i
FROM: PETER SCHOFIELD
DATE : 22 February 1999
EXTN :
ROOM: 25/G
cc: Chancellor
Economic Secretary
Sir Andrew Turnbull
Steve Robson
Harry Bush
Adrian Montague
Robin Fellgett
Adam Sharples
Colin Farthing
Sarah Mullen
Louise Bennett
Ed Balls
Spencer Livermore
BA/POCL: MINUTE TO THE PRIME MINISTER
Issue:
Now that Alistair Darling (letter of 12 February) and
Stephen Byers (letters of 18 and 19 February) have
written in support of the alternative approach
developed by Steve Robson, you and Lord Falconer
need to send a minute to the Prime Minister. I
attach a draft which has been discussed with DSS
and DTI officials.
Immediate.
1
RESTRICTED - POLICY
HMT00000025
HmT00000025
RESTRICTED - POLICY
2. The draft minute runs through the issues, including the costs and the
risks, which were flagged up in the previous two papers that you and Lord
Falconer have received from us.
3. However, in response to one of Stephen Byers’ concerns, this draft
includes a paragraph on the impact of pursuing the new approach on the
timing of the Post Office White Paper. For the reasons set out at the end of
the draft minute, we do not think the White Paper can be published until we
have reached heads of agreement with ICL. This probably means a delay
until well after Easter. We do not see this as a showstopper, but this does *
represent significant slippage - the original commitment was to publish in
January.
PETER SCHOFIELD
PEP
2
RESTRICTED - POLICY
HMT00000025
HmT00000025
CONFIDENTIAL - COMMERCIAL
PRIME MINISTER
HORIZON (BA/POCL AUTOMATION)
Issue
Whether to proceed with an alternative approach to the Horizon project, which has
emerged in discussions between Treasury officials and ICL. There are greater risks,
and potentially additional costs, in pursuing this new approach than if we continued
with the Benefit Payment Card (BPC). It would also almost certainly mean that the
Post Office White Paper would have to be delayed. But on balance, we believe this
new approach fits better with the Government's objectives (as set out in Alan's
minute of 25 January), and by improving the incentives on BA and POCL offers a
better chance of delivering a successful outcome. Both Alistair Darling and Stephen
Byers have written with their support in principle for the new approach.
Action
2 Ifyou agree we should take this approach forward, officials from HMT, DTT
and DSS would work with BA and POCL to develop a detailed specification and
negotiating remit, before opening discussions with ICL. But it would help the
handling with the Post Office if you were prepared to speak to the Chairman or Chief
Executive to let them know of the Government’s decision and make clear the
Government’s commitment to the new option.
HMT00000025
HMT00000025
a
CONFIDENTIAL - COMMERCIAL
Fi Immediate, so that negotiations can begin with ICL as soon as possible.
Background .
4 You agreed (Jeremy Heywood’s letter of 28 January) that Treasury officials
should explore with ICL the options for achieving the Government's objectives for
this project with or without a Benefit Payment Card (BPC).
5; The new approach which emerged from these discussions is set out in detail
at Annex A. It would involve dropping the BPC and moving more quic
payment of benefits by ACT. However, it would also seek to maintain footfall
revenue for the Post Office, since those benefit recipients who currently receive their
benefits via order books would have their benefits paid into simple credit-only
“benefit accounts” run by POCL and only accessible at post offices via a smartcard.
6. It would, as now, be open to any benefit recipient to opt for payment of
benefits by ACT into a conventional bank account (and eighty per cent of benefit
recipients already have access to a bank account). However, we believe that a
significant proportion of benefit recipients will continue to receive their benefits at
post offices - either because they are amongst the twenty per cent who do not
already have a bank account, or because the local post office is more convenient for
them, or just because of inertia. This means that the Post Office should retain a
similar level of footfall as under the BPC. It also ensures that the smartcard has a
large initial circulation, which will help the marketing of smartcard-based services.
HMT00000025
HMT00000025
CONFIDENTIAL - COMMERCIAL
the We have looked at the costs and the risks attached to this new approach
compared with the BPC (see Annex B), with the help of KPMG.
8. On the costs, the initial modelling showed that the BPC appeared to offer +
better value for money than the new approach. There are factors which could
narrow this gap. In particular, KPMG took a cautious view of the impact of the new
approach on Post Office footfall revenue. They also pointed out that the costings
were highly dependant on the timetable - the value for money gap would be
narrowed further if it were possible for DSS to accelerate the start of payments by
ACT, provided POCL had by then installed the infrastructure necessary to support
post office benefit accounts.
e)) However, there are risks from pursuing the new approach which could
increase the costs:
— — ICL may view the change in the project specification as an opportunity to
try to recoup from the public sector a significant proportion of the
development cos
s already sunk in the BPC (perhaps of the order of
£250 million), and in addition the costs of the new development work that
would be needed to meet the requirements of the new approach;
~ there is the risk associated with developing a new infrastructure to support
the post office benefit accounts;
— potentially greater fraud risk borne by the public sector. Under the BPC,
up to £200 million of the fraud risk (eg from stolen cards) was to have
HMT00000025
HmT00000025
CONFIDENTIAL - COMMERCIAL
been borne by ICL. It is likely that ICL will argue that under the
alternative approach, because it will be POCL who would be issuing the
s to holders of post office benefit accounts, it should be the Post
car
Office that bears this risk;
— given the fundamental changes that this new approach will require to the
original contractual basis, there are difficult issues relating to EU
procurement law which would need to be addressed, and a risk of judicial
challenge.
10. On the other hand, the new approach would provide better incentives for the
parties (BA, POCL and ICL) to act together, which would have implications for the
overall success of the project. It would take BA out of the contract altogether - their
focus would instead be on getting their own systems ready for ACT. For POCL, it
provides the infrastructure necessary to help build a business as the supplier of
community financial services. Above all, it offers the best chance of delivering this
project successfully, since the (wo remaining parties to the contract (POCL and ICL)
will have every incentive to see the Horizon infrastructure in pla
soon as
possible.
11. More generally, the new approach should also provide a boost to wider
Government objectives, by bringing on stream more quickly a widely available
smartcard for the delivery of modern government services, and combating financial
exclusion by providing simple bank accounts to all benefit recipients.
12. On balance, we therefore believe that this new approach is worth pursuing.
Alistair Darling and Stephen Byers have both written with their support in principle
HMT00000025
HmT00000025
CONFIDENTIAL - COMMERCIAL
However, Stephen is concerned that taking forward this new approach will delay the
Horizon project. His agreement is on condition that we can keep the BPC option on
the table as a fallback as we take forward work on the new approach.
Way forward
13. So far, the specification and costing of this new approach have been carried out
solely within the Treasury with the assistance of KPMG. The Benefits Agency and
Post Office have not yet been involved. If we are to take it forward, we must
involve them as soon as possible. And we also need to tell ICL that Ministers would
like to take this approach forward.
14. The most difficult handling issue involves the Post Office, particularly in the
wake of the announcement in December that they would be given greater
commercial freedoms. Their Board have already approved the BPC under the
contractual terms offered by ICL before Christmas, so they may feel that they are
how being overruled on a commercial decision. It will be important to make clear
(o senior Post Office management that there are Government-wide reasons why this
approach is preferable to the BPC, and it would help to reinforce this point if you
were able to give this message to the Post Office Chairman or Chief Executive
personally, In addition, if the Post Office are to work seriously on this new
approach, it would be important to make clear to them that the Government’s firm
decision is in favour of the new approach rather than the BPC.
15. Once BA and POCL have been informed, the next stage would be to draw up
with them a detailed specification for the new project and a negotiating remit. While
the detailed negotiations with ICL would have to be taken forward by POCL and
HMT00000025
HmT00000025
CONFIDENTIAL - COMMERCIAL
BA, we suggest that DSS and DTI officials, under the chairmanship of the Treasury,
should oversee and direct the process. In addition (not least to maintain our
negotiating leverage, as well as to meet Stephen's concerns) it would be vital to keep
in play with ICL the options of proceeding with the BPC and of abandoning the
Horizon project altogether. f
16. However, the main urgency is that senior management at ICL are keen to hear
from us whether we want to pursue this new approach. We would like to be able to
give them an indication within the next day or (wo. We would therefore be grateful
for your approval to take matters forward in this way.
Post Office White Paper
17. The Horizon project is central to the future of POCL, and Stephen believes that
the White Paper needs to be as clear and positive as possible about the future of
POCL. At the same time, it would play into the hands of ICL’s negotiators if we
were to announce our new approach on Horizon before we have reached heads of
agreement. We therefore believe that the White Paper will have to be delayed until
we have made sufficient progress with ICL - realistically this means publication after
ster. In the interim, the sub-postmasters will need careful handling to maintain
confidence.
HMT00000025
HmT00000025
CONFIDENTIAL - COMMERCIAL
18. Copies of this minute go to Alistair Darling and Stephen Byers, and to
Sir Richard Wilson.
ALAN MILBURN CHARLES FALCONER
February 1999
HMT00000025
HmT00000025
CONFIDENTIAL - COMMERCIAL
Annex A
BA/POCL AUTOMATION PROJ
This note sets out a possible way forward for the Horizon project, following
discussions between Steve Robson (HMT) and Keith Todd (ICL). These
discussions were conducted on a without prejudice and confidential basis. ICL
accepted that the public sector patties reserve their rights and remedies with respect
to past breaches of contract by ICL Pathway. Both sides agreed that they would not
lake pre-emptive legal action against any other party while discussions are
continuing,
The proposal
2. Under this proposal, the benefit payment card (BPC) would be removed from
the Horizon project. The Benefits Agency (BA) would move directly from the
existing order book system to payment of benefits through ACT.
3. Benefit recipients who currently receive their benefits via order books would
have a Post Office “benefit account” set up for them. BA would pay benefits into
the account via ACT, just as they would into any normal bank account.
4. The Post Office benefit accounts would be simple, credit-only accounts. They
could be accessed at any Post Office using a smartcard, but at least initially they
would not be accessible at banks. The Post Office might enter into a strategic
partnership with a clearing bank to operate these accounts for them. Subject to such
arrangements being put in place by the Post Office, ICL tell us that they could start
HMT00000025
HMT00000025
CONFIDENTIAL - COMMERCIAL
to roll out. a smartcard-based system into Post Offices following national rollout of
the core Horizon system (currently assumed to be Spring 2001).
5. Over time, the Post Office could develop the range of services which were
offered via the smartcard. It could also enter into negotiations with clearing banks
to offer counter facilities, enabling the banks to shrink their branch network.
6. ‘The presumption would be that all order book recipients would be paid via a
Post Office benefit account, and the move from payment via order book to payment
into benefit accounts would be carried out automatically. However, benefit
recipients would at any point be free to ask the BA to pay their benefits via ACT into
a conventional bank account, just as they can now ask for a move from an order
book to ACT.
7. ‘This approach should mean that a large proportion of benefit recipients will
continue to receive their benefits at post offices - either because they do not already
have a bank account, or because a post office is more convenient for them, or just
because of inertia. This means that POCL should have as much certainty over future
footfall as under the BPC. It also ensures that the smartcard has a large initial
circulation, which will help the marketing of smartcard-based services.
Impact on the
Benelits Agency
Advantages:
HMT00000025
HmT00000025
CONFIDENTIAL - COMMERCIAL
- avoids having to develop and implement the IT infrastructure which ts
required specifically for payment of benefits through the benefit payment card
(PAS/CMS/BES), although much of the development work has already been
done:
- should allow earlier move to ACT than under original benefit payment card
project.
Disadvantages:
- delays move away [rom paper-based systems for paying benefits by two years
compared to BPC option.
Advantages:
- should help to maintain footfall revenue;
- gets smartcard in circulation earlier than under BPC option, bringing forward
potential revenues [rom smartcard-based services;
- simplifies Post Office relationship with ICL if Benefits Agency is no longer
a party to the contractual relationship with ICL.
Disadvantages:
HMT00000025
HMT00000025
CONFIDENTIAL - COMMERCIAL
- the cost of managing bank accounts and of the smartcard would be borne by
POCL:
- may not be consistent with POCL’s long term commercial strategy which is
to provide counter services {o all the banks rather than to run its own bank -
accounts,
- restricts Post Office to contracting with ICL. The Post Office have previously
maintained that if they lost the BPC they would wish to re-tender,
ICL
Advantages:
- removes BA from the contract, helping to clarify ICL’s relationship with PO
as its single client.
Disadvantages:
- ICL will want compensation for cancellation of the bpe and payment for the
development of the new facilities;
- ICL would still need to be paid for ongoing bpe work until Heads of
Agreement are signed.
- extra development work required for smartcard-based banking services.
HMT00000025
HmT00000025
CONFIDENTIAL - COMMERCIAL
Other advantages/disadvantages_for Government:
Advantages:
- boost to Modern Government initiative - early introduction of technology to
support Modern Government services;
- boost to social exclusion agenda by providing simple banking facilities to
those currently without bank accounts.
Disadvantages:
- not a clear-cut solution, involves further period of negotiation which may take
three to six months and might result in failure to reach agreement with ICL:
- from a procurement perspective, to cancel a major part of the automation
project and re-shape the remainder involves an inevitable risk of legal
challenge.
Benefit recipients
Advantages:
- unlike the BPC option, provides a bank account to those currently without
one;
- compared to BPC and conventional ACT, provides a smartcard earlier
(although only an advantage if there are attractive smartcard-based services
HMT00000025
HMmT00000025
CONFIDENTIAL - COMMERCIAL
available).
Disadvantages:
- unlike with the BPC, there are no facilities for urgent payments or payment +
to nominated proxy. But BA would find alternative methods of offering these
services.
Taking the proposal forward - negotiation strategy with LCL
12. In order to preserve our leverage on price with ICL (and in the case of the BPC
to meet DTI’s concerns about not putting the whole Horizon project at risk), it
would be vital to keep the Government’s alternative options open (cither to proceed
with the BPC or to pull the whole project) until the commercial terms for this new
approach have been agreed.
13. This will require a careful negotiating strategy with ICL. They will argue that,
as the Government has changed the specification, the Government should pay ICL’s
sunk development costs for the BPC. ICL have already said that they would want
to split the contract in two, with the Horizon infrastructure, excluding the BPC,
rolled out and paid for first, and then a second set of negotiations on the smartcard
and the method of paying benefits. This would need to be avoided, since it would
leave ICL in the driving seat for negotiations on the smartcard.
HMT00000025
HmT00000025
CONFIDENTIAL - COMMERCIAL
Annex B
BA/POCL AUTOMATION: COMPARISON OF THE ALTERNATIVE
OPTION TO CONTINUATION WITH THE BENEFIT PAYMENT CARD
WHAT ARE THE OPTIONS?
Following discussion with ICL, ICL have proposed an alternative option which
would entail:
e — dropping the Benefit Payment Card;
e the Post Office would set up simple “benefit accounts” on the back of the
Horizon infrastructure;
¢ benefits would be paid via the BACS system into these PO accounts, or else
inlo existing accounts:
e the account could be run by a bank on behalf of the Post Office, and would
simply be a vehicle for the cash withdrawal of benefit;
¢ cash withdrawals from a benefit account could only be made from a post office.
We asked KPMG to model scenarios around this alternative option
On the basis of this further work, we believe there are two main options:
Continue with the Benefit Payment Card: option 1
HMT00000025
HmT00000025
CONFIDENTIAL - COMMERCIAL
* project as currently envisaged;
° transition to a smart benefit card from 2001;
¢ — benefits paid by ACT into bank accounts from 2005 when POCL has full >
network banking in place.
Drop the BPC and transfer benefit recipients to a PO benefit account: option
2A
¢ — POCL contract with a banking partner to set up PO benefit accounts accessed
by a smartcard by end 2001;
¢ benefits paid by ACT into PO benefit account from early 2002;
© POCL offer full network banking from 2003.
HMT00000025
HMT00000025
CONFIDENTIAL - COMMERCIAL
THE POTENTIAL RISKS
Comparing the alternative option 2A with option 1:
m the figures are dependent on the outcome and timing of a further round of
negotiations:
— with ICL over the cost of Horizon without the BPC, and the cost of
smartcard. It is far from clear that we could keep ICL to the same NPV
loss negotiated by the public sector parties following Corbett. In
particular, ICL would argue that they should be paid for their sunk
development costs for the BPC, and it is not clear what they would charge
for the smarteard;
— with potential clearing bank partners for POCL to provide benefit
accounts;
~ — with all_the banks to allow POCL to offer network banking servic
(allowing post offices to provide counter services on behalf of banks).
The modelling of the alternative options assumes that POCL is able to
offer network banking services {rom 2003. If this date were to be delayed
it would hit the financial projections for option 2A compared with option
VF
m the fraud risk (e.g. from the use of a stolen card). Under Option 1, this risk is
borne by ICL, and there are various BPC-specific IT systems planned to
manage this (e.g. to verify the identify of the cardholder). Under ACT, fraud
HMT00000025
HmT00000025
CONFIDENTIAL - COMMERCIAL
tisk falls to the banks - in the case of the alternative approach, POCL or
POCL’s banking partner.
the development risk. The development of the BPC is almost complete. In
contrast, the alternative approach requires development of a number of IT +
systems - in particular, the interface with the Horizon platform in Post Offices
and the POCL benefit accounts.
the impact on the network. The reaction of subpostmasters to the removal from
the project of the BPC would have to be carefully managed
HMT00000025
HmT00000025
CONFIDENTIAL - COMM
ERCIAL
WHAT IS ‘THE OVERALL IMPACT ON THE PUBLIC SECTOR?
The following table shows the NPV (1999-2010 discounted to 1999) of options I
and 2A relative to a base case of “business as usual”.
£bn (NPV discounted
at 6% to 1999)
option 1: continue with
BPC
option 2A: PO benefit
account (new
approach) I
net impact on DSS 137 pH
net impact on POCL 0.17 1.29
overall NPV 1.54 121
* the modelling suggests that the alternative option 2A is less beneficial to the
public sector than continuation with the BPC
* whilst the savings to DSS are much larger under the alternative option, the
costs falling on POCL are consequently much greater (reflecting the costs of
setting up and operating the PO benefit account and the loss of BA income).
KPMG also assumed there would be lower footfall under option 2A, though
this looks pessimistic. Assuming no loss of footfall, the NPV for option 2A
would improve by up to £70 million (see refinements section below)
¢ there are significant uncertainties attached to the modelling (although these are
likely to be greater for option 2A than option 1)
* — the NPVs do not fully reflect the risks under each scenario
HMT00000025
HMT00000025
CONFIDENTIAL - COMMERCIAL
RE
NEMENTS TO THE OPTIONS
KPMG also considered sensitivities around the alternative scenarios:
BA
it might be possible to start the ACT transfer more quickly than BA currently
envisage - but we would need to discuss this with BA. In particular, this may
be constrained by BA’s IT systems
¢ would risk moving to ACT faster than PO could implement PO benefit
accounts
BA completes transfer to benefits via ACT more quickly
* completing transfer to ACT over 18 months rather than 3 years would improve
the NPV of option 2A by £130m as DSS savings come on stream more quickly
* — we would need to discuss with BA whether this was operationally possible
* — note this would not close the gap between option I and option 2A
POCL markets its full banking more aggressively under option 2A
NPV on 2A might be improved if PO manage to transfer these customers to full
banking more quickly
HMT00000025
HMT00000025
CONFIDENTIAL - COMMERCIAL
BA/POCL work together to maintain footfall in option 2A at the same level as option
al
* — option 2A entail greater loss of footfall than option 1, since some changes -
implemented by BA (e.y. periodicity) will encourage a larger shift to ACT into
conventional accounts than under option I
* it might be possible for BA and POCL to work together to reduce footfall loss -
possibly saving up to £70m NPV under 2A
* note this would not close the gap between option I and option 2A
POCL offers a more conventional bank account
* — envisages PO bank accounts offering simple bank services. Could be accessed
at points other than the Post Office
* significant risk to PO commercial banking strategy - to become network banker
for all banks - since banks would see this as a competing product
* greater risk to footfall - since accounts can be accessed at other non-post office
locations e.g. ATMs