1. How should the cost and benefits of Private Finance projects be assessed? What discount rate should be used in comparing Private Finance with conventional public procurement? Are current procurement procedures satisfactory? Is enough information disclosed on Private Finance projects fully to assess whether the taxpayer is getting value-for-money?
The cost and benefits of the Private Finance Initiative (PFI) can be assessed on a macro level, as regards the PFI programme as a whole, and on a micro level, on a project-by-project basis.
On a macro level, the question to ask is "where would we be today without the PFI?". In this context, the PFI can be seen as a very successful delivery tool enabling the Government to procure many more infrastructure projects than it would have been able to do in the same time period using traditional procurement methods. In 15 years, the PFI programme has delivered over 510 PFI projects which have now completed construction and are in operation. This has meant that the benefits of new infrastructure have been realised much earlier than by more piecemeal renewal.
A couple of examples. In its 2008 report1, the Treasury noted that in education, the total schools capital investment rose from less than £700 million in 1996-97 to £5.9 billion in 2007-08, and is projected to rise in 2010 to £8.2 billion. On the ground, this means that in Autumn 2009, over 100,000 secondary school students will be returning to learning environments which have benefitted from investment as part of the Building Schools for the Future (BSF) programme alone. Furthermore, a 2008 KPMG report2 indicated a statistically significant correlation between PFI school projects and improved educational outcomes.
In health, the 2008 Treasury report noted3 that between 1996 and 2008 70 PFI (and 23 traditionally) procured hospital schemes were completed, and as at the time of the report another 27 PFI (and seven public capital) projects were under construction, modernising the NHS estate such that only 20% of the NHS estate is pre-1948 (down from 50% in 1997). It is not only a tool for the initial delivery of the assets-PFI is a whole-life cycle solution-it locks in the continued delivery the services, and maintenance of the assets, for the period of the relevant contract (up to 30 years).
How does PFI deliver?: It encourages innovation in project design and delivery; it imports private sector discipline into the delivery of public sector infrastructure and services, resulting in time-certain and cost-certain procurement; it allows the public sector to leverage up limited public sector resources making them go further and enabling delivery of an intensive building programme-the public sector does not have sufficient skilled procurement teams to have been able to achieve the same volume of capex-intensive procurements that have been successfully undertaken since 1997 had it only used traditional procurement methods (which, due to the risks the public sector retains under that method, require more skilled public sector staff during the procurement, construction and operation stages than PFI procurements); by speeding up the delivery of assets, it also accelerates the delivery of the related social benefit (improved healthcare, education, transport); the standardisation of documentation has made possible faster procurement of multiple PFI projects in "commoditised" form; the long-term nature of PFI deals locks in the maintenance of assets, and allows procuring bodies to make long-term plans for the delivery of a service, with limited risk of a change in policy/ Government affecting those plans.
Is another way better?: On a micro level, the question can be phrased "what are the costs and benefits of a particular project as procured under the PFI compared against another method of achieving the same outcome?". Or, put another way, is the project "value for money" (VFM)? The alternative procurement method currently used for making this comparison has been "traditional procurement"-ie the public sector designs and delivers the project and the resultant services, and only contracts to the private sector the construction of the relevant infrastructure. Recent developments in the Treasury's guidance for doing this comparison (the Quantitative Evaluation Tool, QET) have made this assessment much simpler and more transparent, giving a level of independence to the findings.
Room for improvement: First, the cost/benefit analysis is currently routinely only made at the time of the project's procurement-it should also be made periodically during the service delivery period-this would show that operational costs (to the public sector) during a PFI project's life are controlled and continue to offer VFM, and that the quality of the relevant asset is being maintained (compared to traditionally procured assets, where historically chronic under-maintenance and lack of investment led to the severe deterioration of assets). This is reinforced by the payment reductions found in PFI projects-if a project is not being operated and maintained to the required standard, the tariff will be reduced. No such controls on service and asset quality exist for traditionally procured projects.
Use a different comparator: Second, there have been few major traditional procurements of projects that might otherwise be suitable for PFI in recent years and, even for the few there have been, as data on those projects is not routinely collected (the last major collection seems to have been in 2002 for a study Mott MacDonald did for Treasury of 39 traditionally procured and 11 PFI procured projects, which was then used for their 2002 report on optimism bias4) and there is a particular lack of data on the operating phase of traditionally procured assets.5 So the data that is used for the QET is old and therefore perhaps not credible. As such, a better VFM comparator could be a similar PFI project, on which a more straight-forward comparison could be made and for which much more detailed and recent data is available.
Protection of asset condition: Another benefit to be considered in this analysis is the requirement that PFI assets be handed back at the end on the PFI project in a specified condition. This is another means of avoiding the chronic running down through lack of life-cycle maintenance of traditionally procured public assets which was historically the overriding experience for traditionally procured projects.
Discount rate: The discount rate to be applied is a matter for the procuring authority to decide, taking into account the relevant optimism bias assumed in the QET. The underlying discount rate should be applied consistently across all methods of procurement being considered.
Competitive dialogue works: The current procurement procedures are challenging but workable. In the case of competitive dialogue, the theory is good-it should result in more effective competition and therefore better VFM-but the challenge lies in improving its practical application. The Government needs to give procuring bodies more direction on how the successive stages of the competitive dialogue procedure should be run, particularly the appropriate level of detail that the competitive dialogue procedures need to be invoked for at each stage of the bid. Currently, procuring bodies are erring on the side of caution and invoking them at every stage, which is pushing the bidding costs up for limited benefit. For example, what is the benefit of two bidders doing two sets of title investigations, incurring one set of unnecessary legal costs, when one set could be done and shared? If there was more guidance on this point, then there could be significant cost and time savings. As such guidance would reduce the complexity of the procurement process, it may also reduce the risk of challenges to the tender award.
Not enough information: There is plenty of information on PFI projects available-on both the construction and operation phases-to assess VFM. But if traditional procurement is used as the VFM comparator, there is not enough information on conventionally procured projects, and the limited information that exists is too old, to make a meaningful comparison-see comments above.