2.16 Most PFI contracts allow an annual increase to the contract price to keep pace with retail inflation. Recently however, wage costs have risen above retail inflation due to, for example, increases in the national minimum wage. The average earnings index4 rose by 4.0 per cent during 2000-2005 compared with 2.5 per cent for the RPI.5 The introduction of minimum pay rates under the Department of Health's Agenda for Change, applicable from September 2005, will further affect wage costs in NHS service contracts. As shown in Figure 11, prices following value testing had increased in four of the seven building projects that had completed the process, with two other projects securing small price decreases. Although benchmarking and market testing had allowed the parties to determine a current market price for the services, this had usually been followed by a negotiation phase during which the public sector had, in some cases, reduced the scope of the services being provided to limit the scale of any price increase.
2.17 Based on information supplied to KPMG by the private sector6, most of the respondents reported that they were making profits on their services, 17 per cent were not and 38 per cent were not performing to operator expectations of profitability. This could suggest that some initial contracts for these services may have been priced too competitively at below normal market rates so that when costs increased, the contracts became unprofitable. In these circumstances it is still feasible that investors in the project (the SPV - Figure 6), can make positive returns on their investment if they have transferred the risk of rising costs to their subcontractors providing facilities services. If the contractor is also an investor, any losses as a contractor may be offset by their returns as an investor in the SPV.
2.18 The final price increases from these outcomes were mostly in the range -2 per cent to +6 per cent, although in the Debden Park School project there was a final price increase of 14 per cent. The final price increases were agreed following negotiations, initiated by the authorities, in response to higher price changes which were initially requested by the private sector.
2.19 The recent experience of rising facilities management costs above the annual increase in contract price for general inflation has brought increased pressure to bear on the private sector to, either make economies in the provision of these services, or to seek redress through price increases following benchmarking or market testing. The initial results of the seven value testing exercises of the soft FM projects we examined showed that, based on existing service levels and before taking account of Agenda for Change in the NHS, the private sector proposed price changes of between -1 per cent and +26 per cent with most seeking changes in the range -1 to +1 9 per cent (Figure 15 on pages 24 to 27). These proposed price changes were additional to the annual price changes for general inflation allowed by the PFI contracts.
2.20 This upward pressure on prices has influenced the authorities' approach to dealing with the results of the value testing process. In all the cases we examined, the public sector instigated some form of negotiation about the initial price change proposed, as is likely to be needed as part of the benchmarking or market testing process, which led to a reduction in the price increase that had been proposed. As part of these negotiations public sector teams saw benchmarking/market testing as an opportunity for the authority to consider what services were now required and at what level of specification. As a contract progresses, demand for services evolve, sometimes as a result of government policy (such as the initiative for cleaner hospitals) and sometimes in recognition that some services can be operated in a more efficient way. The authorities we examined considered these specification issues together with their wish to keep the contract price affordable following the value testing.
2.21 Figure 15 on pages 24 to 27 shows how the final price changed following these negotiations and how much the change was accountable for by changes in specification. This shows that, for the soft FM projects examined, the final price changes were lower than the private sector had initially proposed but there had been service specification changes to accommodate this. For the soft FM projects examined:
Final price changes1 were: | -2% to +14% Mostly -2% to +6% |
After negotiations from the initial price changes proposed: | -1% to +26% Mostly -1% to +19% |
Note 1: excluding service enhancements requested at the time of the value testing and the effect of Agenda for Change in the NHS.
These projects were the first PFI building projects to use value testing processes and the price changes reflect cost changes in the market for facilities management services up to 2006. The resulting price changes are not, therefore, indicative of the price changes that may arise in future uses of value testing in PFI building projects.
2.22 In the final price changes, only Debden Park High School, at 14 per cent, fell outside the range -2 per cent to +6 per cent. In the case of Debden Park High School, the like-for-like price increase was primarily due to:
■ an uplift to the catering price which was previously being run at a loss by the contractor; and
■ a large increase in the price for grounds maintenance (which was assessed by the Local Authority against similar schools).
However, two service enhancements brought the final agreed price rise to 25 per cent:
■ a need for catering to reach higher standards due to government initiatives regarding healthy eating; and
■ the appointment of a new caretaker (as a result of a decision to improve the service performance).
2.23 The price changes set out above exclude the effect on the cost of services supplied to hospitals of the Agenda for Change7, which was still being discussed by the NHS Trusts we visited and their contractors at the time of our fieldwork and would have happened regardless of the value testing process. Subsequent to this, one NHS Trust, Queen Elizabeth Hospital, Greenwich, has adjusted its PFI contract price for the effect of Agenda for Change salary increases. The result has been to increase the price adjustment following its market test from 6 per cent to 37 per cent (attributable to £1.2 million for the Agenda for Change effects), although further price adjustments increased the final adjusted price change to 40 per cent. Other price changes in the health projects excluded the impact of further salary increases arising from the Agenda for Change, which has applied to the private sector since October 2005 but the handling of which was still being discussed by authorities at the time of our study.
2.24 Where authorities had reduced their service specification so that the contract remained affordable following the benchmarking or market testing, they considered that this had been effected by identifying efficiencies, which would not affect the service to the user of the facilities. The changes reflected aspects of the specification which they considered to be over specified or which could be delivered more efficiently. For example, following its benchmarking exercise, Darent Valley Hospital identified that office cleaning standards could be achieved satisfactorily with a reduced cleaning regime and that it was not necessary to have two dedicated porters for operating theatres. Queen Elizabeth Hospital, Greenwich, informed us that some enhancements to the service which were being considered at the time of the market testing were not taken up in order to keep the cost affordable. Figure 15 overleaf provides some initial information on service performance following the value testing exercises but it is too early to judge with certainty whether the authorities' expectation, that service specification changes following benchmarking or market testing would not affect service delivery to the public, will be realised.
2.25 The prices that authorities pay for facilities services following benchmarking or market testing may also be different from those that had applied at contract letting due to changes in operational requirements since the contract was let. For example, at Darent Valley Hospital, reductions in FM service costs arose because the Trust had taken a separate decision, as part of its clinical strategy, to close a number of beds. This variation in service requirements occurred at the time of the benchmarking, but the Trust had been planning the change previously in conjunction with the local Primary Care Trusts in order to improve the efficiency and effectiveness within the hospital as part of a NHS-wide policy to increase efficiency. The Trust does not expect the bed reductions to affect patient care negatively as this will be balanced by improving day case rates and reducing the length of patient stay in line with current NHS practice.
2.26 The price changes that have arisen from these value testing exercises should not be viewed in isolation but are part of the overall cost of procuring facilities services to the standards specified in the PFI contracts over an extended period. The Treasury requires project teams to make a value for money decision on whether to include or exclude these services from the contract before embarking on a PFI procurement. We are not, however, aware of any systematic overall comparison to date between the cost and quality experiences of facilities services procured under the PFI with conventional outsourcing. There are difficulties in making these comparisons as the conventional examples may not be comparable to the PFI deals in terms of the required services or standard of performance. There is also limited experience to date of the price changes arising from using the value testing arrangements in PFI contracts. We consulted with our consultants Arup, the Institute of Facilities Management, the Business Services Association, the Office for National Statistics (ONS) and the Treasury and found that there is limited public data available on non-PFI procurement costs of these services. ONS has produced experimental data on the cost of catering and cleaning starting from 2000 but this data is not official ONS data due to its experimental classification.
[Access 15 Changes in prices following benchmarking/market testing (BM/MT) – PDF]
16 | Issues affecting whether projects were likely to achieve value for money from the value testing process | ||||||||||
| Darent Valley Hospital | University Hospital of North Durham | Defence Fixed Telecommunications Service | Debden Park High School | St John's House, Bootle | Sussex Partnership NHS Trust | Foreign and Commonwealth Office Telecommunications Network | National Savings and Investments
| Hereford and Worcester Magistrates' Courts | ||
| Did the PFI project have a comprehensive collection of data? | Yes Public sector adviser collected an extensive range of comparative data. | Uncertain The contractor provided anonymised data and the comparator data was based on a limited number of NHS Trusts | Yes Comprehensive market data is collected on a three monthly rolling basis, but adjustments are needed to compare MOD data to non-military communications systems given the MOD's additional security requirements. | Uncertain There were some concerns expressed by the Authority that the contractor had not always used appropriate comparators. However, there was reluctance on the part of other authorities to provide benchmarking data since they saw this as confidential information.
| Uncertain It was difficult to get agreed comparative data. | Uncertain It was difficult to find other mental Health PFI units as a comparator. | Uncertain There was limited data available to act as a comparator due to the specialism of the services and the required flexibility demanded by the Foreign and Commonwealth Office (FCO). | Yes However the speciality of the services involved makes it difficult to compare. In addition, benchmarking is on efficiencies rather than pure cost data. | Uncertain The scale and scope of the data collected is not clear. | |
| Did the benchmarking process offer an effective means of price adjustment? | Yes The data collected was effectively analysed and used in negotiations with the private sector. | Uncertain There were data issues regarding regional variations and outliers which meant that the benchmarking process may not have been effective as possible. | Uncertain The benchmarking process was undertaken under a previous contractual arrangement, which although allowing annual price adjustments, was limited to two challenges a year on the seven service groups. This restriction has been removed in the renegotiated contract. | Yes An outcome was agreed and there was full visibility and dialogue between the parties. | Yes Although the benchmarking exercise took 25 months to resolve, the price increases suggested by the contractor were not accepted. | No The parties had to resort to market testing. | Uncertain FCO considered market testing at one point but they believed the contractor remained the best value credible supplier. However they consider that there are now a number of credible potential suppliers. | Yes The internal benchmark studies undertaken jointly between NS&I and its partner, Siemens, are effective as they are part of an annual planning cycle that targets those areas to be benchmarked with emphasis on the delivery of benefits for both parties in terms of improved efficiencies and better services to customers. | Uncertain There is only a downward price adjustment allowed on the first benchmarking which may have had a detrimental effect on the benchmarking process since the incumbent is not incentivised to engage fully with the process. | |
| Outcomes |
|
|
|
|
|
|
|
|
| |
| 1. Price increase, decrease or stayed the same | 1. There was a 2.4% reduction in the unitary charge for soft FM services. | 1. A soft FM price increase was offset by a hard FM decrease to give a small net increase in price (1.48%). | 1. There was a 37% decrease in the price of the benchmarked service. | 1. There was a 14% price increase in the services. The price increase was largely accountable by an uplift to a service which was being run at a loss and an increase in grounds maintenance which was assessed by the Authority against similar schools. | 1. There was no change in the price of the services. | See market testing overleaf. | 1. The unitary Charge has consistently gone down in line with market trends. | The benchmarking process is still in progress (Figure 10 and 15). | The benchmarking process is still in progress (Paragraph 2.13). | |
| 2. Service performance since the value testing | 2. According to the Authority there have been no adverse impact on patient services although some reduction in the service specification has occured in areas which the Trust considered over specified or which could be delivered more efficiently. | 2. Service performance has been good. A small service saving was made by bringing part of the IT service in-house. | 2. The service has been improved since the benchmarking. | 2. The School has expressed some concerns regarding cleaning standards. Service enhancements were also negotiated at the time of the benchmarking.
| 2. There were no changes to the services.
| See market testing overleaf. | 2. Service performance has been unchanged. |
|
| |
| National Audit Office Value for Money assessment Is it likely that value for money had been achieved through the benchmarking? | Yes There was a reduction in the unitary charge for soft Fm services but according to the Authority no adverse impact on patient services. | Uncertain Lessons learnt from this first process have been incorporated into the second process which has commenced. This is designed to ensure better value for money in the outcome. | Yes The result was a price reduction with an enhanced service. The use of advisers who review service prices on a quarterly basis has given MOD added assurance that services are being scrutinised rigorously. | Uncertain The value test resulted in a 25% price increase. The school has expressed some concerns regarding cleaning standards. There was a significant price increase and a possible reduction in service standards | Yes The Authority successfully opposed a 16% price increase. | See market testing. | Yes The result was a price reduction with no change in service performance. | Uncertain |
| |
Norfolk and norwich University Hospital | Queen Elizabeth Hospital, Greenwich | Sussex Partnership NHS Trust | |||||||||
Did the project attract sufficient market interest? | Yes | Yes | Yes | ||||||||
| There were 16 expressions of interest which reduced to six with a final reduction to three in the process. | Three alternative suppliers bid. | Although initially two bidders withdrew from the process two new suppliers bid in the final stage of the process. | ||||||||
Did the market testing process offer an effective means of price adjustment? | Yes An outcome was achieved. | Yes An outcome was achieved. | Uncertain The Authority decided to bring the services in-house. | ||||||||
Outcomes |
|
|
| ||||||||
1. Price increase, decrease or stayed the same | 1. There was a 2.17% reduction in the unitary charge for soft Fm services. | 1. There was a price increase of 6%, the highest of the hospital projects examined. The Authority believes that this may have been greater under a benchmarking process. | 1. There was a price increase of 5.7% | ||||||||
2. Service performance since the value testing | 2. Service performance has been good. | 2. Service performance has improved. Service enhancements were also negatiated at the time of the benchmarking. | 2. Service performance is as planned. | ||||||||
National Audit Office Value for Money assessment |
|
|
| ||||||||
Was value for money likely to have been achieved through the market testing? | Yes The result was a price reduction with no detrimental change in the level of service performance. | Uncertain There was a price increase higher than the other hospitals examined. | Uncertain The Authority decided to bring the services in-house as the revised cost of the services was lower than that obtained through benchmarking. However, this meant that the Authority would hold the risk associated with the delivery of the service. This risk was not formally calculated since the Authority considered that the difference in the risk was at worst neutral and at best in the Trust's favour. The Authority also considered that the risk of completing the market testing process was that the Trust could then be committed to an unacceptable price increase. | ||||||||
National Audit Office Value for Money assessment – see pages 30-31 | |||||||||||
Source: Project reviews conducted by the National Audit Office and ARUP | |||||||||||
NOTE | |||||||||||
We considered the value for money of the completed value tests by reference to the resulting changes to the price and service specification, and the effectiveness of the value testing. Uncertainty was due to various factors including whether the best price had been secured where there had been price increases, weaknesses in the comparator data and authorities identifying that they would make changes to their value testing processes in future to achieve better outcomes. | |||||||||||
__________________________________________________________________________________________________
4 Office of National Statistics: Average Earnings Index (seasonally adjusted).
5 Office of National Statistics: Retail Prices Index (all items).
6 KPMG review of PFI projects 'Effectiveness of operational contracts in PFI 2007'. Standard & Poor's PFI report 2006 also referred to losses being incurred on some PFI facility management contracts.
7 Agenda for Change, see footnote 2, page 6.