INTRODUCTION
AND
SUMMARY
OF
CONCLUSIONS
AND
RECOMMENDATIONS
1. The Private Finance Initiative (PFI) contracts
for Bridgend and Fazakerley prisons were the first PFI prisons
contracts to be let and were also the first contracts involving
the construction of buildings to be awarded under the PFI.[1]
2. The Prison Service let the contract for the 800
prisoner place prison at Bridgend, South Wales to a consortium
led by Securicor and Costain in January 1996, shortly after the
award of the contract for a 600 prisoner place prison at Fazakerley,
Merseyside to a consortium led by Group 4 and Tarmac in December
1995. The Service estimated that the Bridgend contract would cost
(in 1995 prices) £266 million and the Fazakerley contract
£247 million.[2]
3. The contracts require the private sector partners
to design, construct and finance the new prisons and to operate
them for a period of 25 years with the buildings being transferred
to the public sector at the end of the period. The contractors
will be responsible for all custodial services and other services
such as catering, prisoner education and onsite medical facilities.[3]
4. On the basis of a report by the Comptroller and
Auditor General, the Committee took evidence from the Prison Service
on how the costs and benefits of the chosen PFI solutions compared
with alternative forms of procurement, why the Service decided
against letting both contracts to a single bidder, and how they
managed the costs of their advisers. We subsequently obtained
written evidence from the Service about operational difficulties
experienced after the opening of Bridgend Prison (now called Parc
Prison) and the costs of using police cells as temporary accommodation.
5. We recognise that both prisons came into operation
ahead of target, twice as quickly as prisons built under traditional
procurement, and that the Service estimated that the Bridgend
contract offered savings (in 1995 prices) of £53 million
compared with the public sector alternative of building the prison
under traditional procurement and then contracting out its operations.
We note that in future PFI procurements the Service intend to
seek a greater transfer of risk, further reduction in unit costs
and cost effective innovation.
The Committee is, however,
concerned that there have been significant operational problems
at the Bridgend prison where Securicor are the operator. The Service
were aware of Securicor's lack of experience in running prisons
and we are concerned that the Service did not identify and address
the shortcomings in Securicor's operating proposals before letting
the Bridgend contract. The Prison Service must now ensure that
they put into effect the lessons they have learned from these
operational problems.
We also note that the Fazakerley
contract offers only relatively small savings of £1 million
(in 1995 prices) compared with the public sector alternative of
traditional building procurement and contracted out operations.
The Prison Service decided against awarding both contracts to
Securicor/Costain even though they submitted the lowest contract
price for Fazakerley. The Service gave us three reasons for their
decision: increased risk of delay if one contractor was used;
they wished to create a market in PFI prisons; and they wanted
to avoid using expensive police cells if the prisons were delayed.
But they failed to quantify the savings foregone and to consider
ways of securing the considerable potential benefits from using
one lead contractor on both projects. One way forward might have
been for the Service to have explored with Securicor/Costain the
possibility of their sub-contracting part of the work or using
at Fazakerley Costain's innovative design proposals for Bridgend.
The Prison Service appointed
their financial advisers, Lazards, to whom they paid £333,000
(more than double their estimate of £145,000) without competition.
We criticise the Service for ignoring recommendations from our
predecessors, and guidance from the Treasury, that advisers should
be appointed through full and open competition.
6. Our detailed conclusions and recommendations are:
On the Prison Service's award of the contracts
(i) The Committee
is concerned that there have been a range of operational problems
at the Bridgend prison since it opened where Securicor are the
operator. We note that the Service had been aware, before letting
the Bridgend contract, that Securicor had no experience of running
prisons and would need to take on board expertise at management
level. This was a reason why the Service would not award both
contracts to the Securicor/Costain consortium despite the fact
that they had bid the lowest price for both prisons. We are very
surprised that, despite these concerns, the Service did not examine
more closely Securicor's proposed staffing and management proposals
for the Bridgend prison. Given their extensive experience of operating
prisons under public sector management the Service should have
been well placed to assess prison staffing and management plans
(paragraph 33).
(ii) We note that the Service have incurred
costs arising from the operational problems at Bridgend prison.
They relate to the cost of providing support staff to the prison
(which they estimate cost £18,400) and other costs such as
senior management time in dealing with the problems. We note that
on account of these problems the Service have deducted £54,782
from payments due to Securicor/Costain, that they are seeking
to make a further deduction of £51,915, and that they expect
Securicor/Costain to bear the full cost of any additional staff
which Securicor supply to the prison. We are concerned, however,
that the penalties for non-compliance with performance measures
are limited to 5 per cent of the annual contract price. We also
note that the Service do not intend that prisons providing support
staff to other prisons should be reimbursed for this service.
We think this unsatisfactory. In our view prisons should meet
the costs that they incur, otherwise it will not be clear what
each prison is costing (paragraph 34).
(iii) We are also concerned that, although
each of the three leading bidders submitted tenders for both of
the contracts on offer, and one of them (Securicor/Costain) said
that they would cut their bids if they were offered both contracts,
the Prison Service did not explore with the bidders the extent
of such savings to the taxpayer that might have been available
(paragraph 35).
(iv) The Service gave us three reasons for
not being willing to negotiate the award of both contracts to
one bidder: that having one contractor responsible for both might
increase the risk of delay in delivering the service; that they
wanted to create a market in PFI prisons; and that if the prisons
were late in opening they would have needed to use expensive police
cells (paragraph 36).
(v) As regards the first, all contracts
carry the risk that the contractor will not deliver the service
required. But the Service were in no position to take an informed
decision without carefully exploring how the particular risks
of awarding both contracts to one bidder might have been addressed,
and what the potential savings might have been. The Service disputed
that, based on the savings offered by Securicor/Costain, the savings
could have amounted to £65 million but agreed that they
would have been about £30 million. We recognise that
the Service were concerned about allowing Securicor to operate
both prisons because of their lack of experience. It is possible
nevertheless that savings might have been secured had the Service
addressed the shortcomings in Securicor's operational plans and/or
negotiated whether Costain could also have worked with another
operator. Nobody will ever know for certain the amount of the
savings that could have been secured because the Service did not
pursue these matters with Securicor/Costain and the other bidders
(paragraph 37).
(vi) As regards creating a market in PFI
prisons, we recognise that these were the first prison contracts
to be negotiated. But the number of bidders for these and subsequent
contracts shows that there is active competition for this business.
Furthermore, PFI contracts are always likely to be more expensive
because of the higher financing costs paid by the private sector,
and passed on to the taxpayer through the contract, unless they
can be offset by innovation in design and delivery. If the market
sees that contracts are in fact being spread around that will
take the edge off the stimulus to produce innovative bids and
the taxpayer is likely to end up with an unsatisfactory deal (paragraph
38).
(vii) As regards the cost of alternative
accommodation, we are concerned that on average it costs £300
for each night a prisoner has to be accommodated in a police cell,
and more in Wales, and that these costs represent net additional
costs to the taxpayer and mainly relate to staff costs, often
at premium rates. We are surprised at the high level of these
charges (paragraph 39).
On whether better solutions could have been achieved
(viii) We regard it as
unsatisfactory that the Service failed to obtain from the contractors
compensation arrangements for possible delay in delivery or operational
problems which would match their assessment of police cell accommodation
costs. We note that the Service expect to be able to transfer
more of this risk to contractors as the PFI process develops,
and we look to them to do so (paragraph 55).
(ix) We are concerned that the Prison Service
were unable to use at Fazakerley the cost saving approach offered
by Securicor/Costain. We regret that the Service did not examine
with Securicor/Costain whether this would have been possible,
for example, through sub-contracting, while spreading the risks
which might have arisen through using the same contractor on both
projects (paragraph 56).
(x) The arrangement whereby Skanska International
Building AB, part of the Securicor/Costain consortium, assisted
Costain on the Bridgend building work was one way of spreading
risk. We recognise that bidders will wish to protect their intellectual
property in any innovative proposals they put forward. But in
our view this need not prevent the Service seeking to engage bidders
in innovative ways of sub-contracting work in order to lay off
the risk of using one contractor on a group of projects (paragraph
57).
(xi) Alternatively, the Service could have
considered awarding separate building and operating contracts
which they have previously been able to use effectively with good
savings. In this connection, we note that they have been considering
whether, in future, other PFI variants can be used such as letting
a contract for the design, construction and financing of a prison,
with the prison being managed by the public sector. We note that
their recent review of this issue concluded that separate building
and operating contracts do not offer value for money compared
with contracts which combine these aspects and the financing of
the project. Nevertheless, given their concerns about Securicor
as an operator at both prisons, a separate building contract might
have enabled them to make use of Costain's cost saving design
at both Fazakerley and Bridgend (paragraph 58).
(xii) We recommend that the Service should
include in their option appraisals consideration of these various
methods of procurement (paragraph 59).
On the appointment of advisers and their costs
(xiii) We find it inexplicable
that, despite numerous previous recommendations from our predecessors,
and guidance from the Treasury, the Service appointed their financial
advisers (Lazards) without a competition. Advisers should be appointed
through full and open competition in all save the most exceptional
circumstances (which did not apply here as these were large procurements
being planned over an extended period). This is particularly critical
where departments are engaged in new undertakings such as a PFI
procurement. We look to the Prison Service to ensure that in future
their advisers are appointed through competition (paragraph 66).
(xiv) We note that legal fees were the largest
component of the advisory costs incurred by the Service. We are
concerned that the Prison Service set an initial estimate for
no more than four months work from their main legal advisers but
that, in the event, their fees of £634,000 exceeded the initial
estimate by more than three times. We are also concerned that
the total cost of legal advice was £821,000 which exceeded
the initial estimate by more than four times. We note the Service
had difficulty in estimating the costs of legal work in these
first PFI prison projects and that they have learned from this
initial experience. We look to them to budget for, and control,
the costs of legal advice in future PFI procurements (paragraph
67).
1 
