Finance Committee, 06 Nov 2001

Meeting date: Tuesday, November 6, 2001

Official Report 210KB pdf

---

Private Finance Initiative/Public-Private Partnerships Inquiry

We move into public session for agenda item 2, which is the first evidence-taking session of our PFI/PPP inquiry. I am pleased to welcome our first witnesses: Amanda McIntyre and Matthew Farrow from the Confederation of British Industry and the Confederation of British Industry Scotland respectively; and Professor Michael Mumford from the University of Lancaster. They have given us submissions. I intend to invite one of you from the CBI to make an opening statement, then Professor Mumford to do likewise. I understand, Professor Mumford, that you also have a presentation for us.

I have some slides, if that will help the discussion.

We all have copies of the slides. I believe that Ms McIntyre and Mr Farrow have copies as well.

We have had an apology from Donald Gorrie, who has another meeting this morning. I ask, as I always do, that anyone who has a mobile phone ensures that it is off and that pagers are switched to buzz.

I invite Mr Farrow to make opening remarks on behalf of the CBI.

Good morning, committee members. I will make a short opening statement, which expands on some of the issues in our written evidence. Amanda McIntyre, who is based in London and has UK-wide responsibility for all our work on modernising government, will answer all the difficult questions—that is my intention, anyway. I will pick out some key themes. I am aware that this is your first evidence session and that you are interested in some of the broad principles.

CBI Scotland has a broad membership, so our interest in PFI is fairly wide. We have firms that are involved in contracts—for example financial services and construction firms—but the wider business community also has a collective interest in the efficient delivery of public services. We have two broad interests. First, business is a stakeholder in public services. Business and its employees consume a lot of public services. That is a particular issue here, where we are always trying to attract talent to business in Scotland and to retain it. Efficient and effective public services are part of the package of attracting people to live and work in Scotland.

Secondly, in a broader sense, from the economic point of view, business is keen that the tax and spend ratio does not rise too far and that it is not out of synch with our competitor countries. At the moment, the proportion of gross domestic product that goes on public spending is drifting up through 40 per cent, which we regard as a threshold. That means that government in the UK and in Scotland will be under pressure to ensure that the public have high-quality services that do not cost more and more money.

The debates of recent years show that the public are not particularly interested in the amounts of money that are spent on services. They are not particularly impressed by the large sums of money that are spent on health, education and so forth—what they want, and what business wants, is effective delivery of high-quality services. In that context, PFI is a useful tool for the Government and the Executive. We do not think that it is a panacea and we do not think that it has the answer to all the questions, but PFI is particularly good at addressing some of the weaknesses of traditional procurement methods.

We pick out three or four weaknesses in particular. One is an obvious point: the political vulnerability of capital budgets and, indeed, maintenance budgets in a political system with annualised accounting—politicians looking to re-election and so forth. For example, if you examine the Scottish transport system—a subject in which business has a particular interest—there are huge fall-offs in spending not just on new road construction, but on maintenance budgets. Local roads are in a terrible state of disrepair. The Convention of Scottish Local Authorities estimates that there is a backlog of £1 billion. In the water industry, which is a political issue in Scotland right now, a lot of the problems relate to decades of underinvestment. Investment obviously is politically vulnerable.

We have often seen public sector-run projects involved in cost and time overruns. That is not good for anyone. Because of the short-term approach in the public sector, with traditional procurement methods, the sponsors of projects—whether they be ministers or departments—have a vested interest in underestimating the costs, because in those circumstances it is easier to get political backing, finance or Treasury approval, which leads to less interest in the whole-life cost of projects.

The benefit of PFI is that it requires politicians and the Executive to consider the cost of assets and service delivery over longer periods. That is especially important. PFI gives the Executive the option of transferring risk to where it is best placed, which can be the private sector. One could say that the high cost of borrowing under PFI is the premium for transferring that risk.

PFI is a useful option for politicians: it enables the Executive to access private sector expertise. Public spending can be very lumpy: there can be years of high public spending on capital projects followed by years when the spending is much lower. If there has not been a big project—a road project or a hospital project—for years or even decades, it can be difficult for the public sector to have all the skills in place to manage such a project. PFI usefully enables the public sector to access and bundle together private sector expertise.

There has been a lot of interest, especially in the press, in the accounting and ownership issues around PFI. We feel that PFI gives the Executive the option of transferring spending on long-term projects from the capital side of the equation to the current spending side. That is not good or bad in itself; it is simply a useful tool that can be the sensible choice in particular cases. From the media's point of view, and from some politicians' point of view, it is true to say that that ability is the main benefit of PFI—the ability to get borrowing off the balance sheet and to spread costs over a longer period. As I say, that is a useful option, but from a business point of view it is only one of the many advantages of PFI and probably not the most important.

The effect of PFI on ownership is, again, neither good nor bad in itself. The business community and, I think, the wider community, want an Executive that is good at providing collective goods and public services in the way that people want and to the quality that people want. Whether that Executive, that public sector, has a huge book of assets that may or may not have values that change over time is less important. That is true for the business community and, I suspect, more widely. Other issues arise and I have no doubt that we will pick them up in our discussion. As I have said, PFI is not a panacea, it is a useful tool. It is an evolving and developing tool for the Executive to work with.

I will talk about two or three of the issues that we think need to be considered to improve the process. The public perception of PFI is important. That is why we welcome this committee's inquiry. Much of the public debate about PFI generates more heat than light and gets caught up in one or two headline issues about accounting and ownership. Much of the more complex detail on the way in which PFI can improve standards of delivery and benefit the consumers of public services tends to get lost. Public procurement stuff has not, in the past, been of great interest to many people, but there is need for a more informed public debate.

This is a learning process. It represents a big change in the way of doing things. The private sector and the public sector are both learning. A lot of good work is being done—by Partnerships UK, for example—to try to spread good practice. There has been good work on the standardisation of contracts. We must ensure that there are ways for the public and private sectors to keep on learning.

Staff issues are important. We were never huge fans of compulsory competitive tendering, partly because of the staff issues that tended to go along with it. In retrospect, it is fair to say that a lot of CCT tended to be bargain basement shopping—going for the lowest price rather than the best value for money. The losers were the consumers, because service quality suffered. The conditions of the work force often suffered too. At UK level, we have done a lot of work with the Trades Union Congress on the Transfer of Undertakings (Protection of Employment) Regulations, which deal with the ways in which staff are transferred or seconded between the public and private sectors. We want to ensure that there is a win-win situation. When providing public services, business must have well trained and motivated employees to deliver the contract to the required level—otherwise, business will be penalised. Business therefore has a vested interest in ensuring that staff issues are considered carefully. The CBI is especially interested in them. I am sure that many of these issues will be picked up in discussion.

Although I initially provided eight slides, I have cut the number down to seven by removing the one headed "Ownership does not matter" because it would take too long to explain the issues.

The whole issue of PFI and PPPs—I will not refer separately to PPPs as they are almost the same thing as PFI—is part of an agenda that was introduced in the early 1980s and included privatisation, executive agencies and compulsory competitive tendering. As a result, PFI first appeared in the context of getting rid of public functions. As Matthew Farrow pointed out, it tends to be seen as a way of reducing costs, but there is rather more to it than that.

The first point to note about PFI is outsourcing, which is just as common in the private sector as it is in the public sector and gets rid of the functions that the firm or public sector body did not do particularly well itself. The principle is that, if an organisation concentrates on what it does best and outsources or buys in the functions that it does not do particularly well, it will become more efficient, knowledgeable and skilled. That requires matching responsibility with authority and so on.

However, there are always difficulties in knowing what to outsource, whether by PFI or through a private firm. We are comparing alternatives, one of which—the one that is not chosen—is bound to be hypothetical. The problem is in preparing a public sector comparator, or PSC. I am sorry to tell the committee that such acronyms crop up everywhere when we talk about PFI. It is very often difficult to assess the quality of what we receive under a contract and to compare that with the quality that we would have received if we had chosen the public supply of the same services. As a result, there tends to be a concentration on cost control and cost reduction. Although the quality of service is often just as important, it is often very difficult to specify that precisely and to assess its performance. PFI deals are usually very long term. For example, the shortest are seven-year IT deals, whereas the longest can run for 30 years and can then be renewed for a further 30 years.

I am concerned with the principles that underlie PPP. It is worth taking a moment to consider the theoretical background. The theoretical base is largely the area of transaction cost economics, which developed in about 1980 and make a distinction between three basic types of contract.

First, there are day-to-day purchases. When we buy Band-aids from a shop, we know what we are buying, the price of the goods and that there are a number of other places where we can buy them. There is no great problem with such purchases and they form the area that much conventional economic theory addresses. A lot of market economics assumes that a product and price can be defined and that competition will automatically produce better value for the purchaser and society at large.

Secondly, we have the purchase of core assets, which are what a business organisation needs to own to carry out its operations. Over the past 20 years, there has been much discussion about what a firm needs to do for itself and what it can buy in from outside. Transaction cost economics theory largely explains that issue in terms of how specific the assets need to be.

The assets do not have to be physical equipment and plant: they can be human-embodied assets such as skills and organisational design. The firm must consider what it needs to do for itself and then what it can buy through long-term contracts from outside—not day-to-day purchases, but commitments to long-term contracts.

When goods or services are to be supplied over a long period, we are typically talking about firm-to-firm deals—the seller would not supply the goods or services without the security of knowing that there was going to be sustained demand over a period, as such contracts require specialised facilities. The purchaser would not be able to obtain suppliers of the goods and services without a long-term contract, because they would not find a supplier who was prepared to supply the skills or quantities that were required. So, we are talking about long-term contracts.

Long-term contracts are all the more complicated if there are some rather nasty technical difficulties. I list them on the slide, but I will refer to only some of them. Long-term contracts are more difficult, first, when there is bounded rationality and opportunism. Opportunism means that there is a risk—not a certainty, nor even an expectation—that people may cheat under the terms of the contract. Secondly, small-numbers bargaining may mean that there may not be many plausible suppliers or customers. We are not talking about highly competitive markets, as there are not enough parties on either side. Thirdly, there may be information asymmetry, which means that some parties will have information that the other parties will not have. That information may be about the specific assets or about how the parties will work with the knowledge that they have. Without information asymmetry, there is no scope for opportunism—there is no scope for cheating if people know precisely what is being done.

Fourthly, there are specific assets, which are assets that have a limited range of uses or irreversible commitments. The term applies to human or non-human assets that have no or few other uses. The more investment in specific assets that is required, the more serious the contracting problems become. Finally, infrequent deals are deals that will not happen every few days, weeks or months. We may be talking about contracts under PFI that last for as long as 60 years.

Those features are described in rather more detail in the bundle of papers that I have supplied in case members want further information. I have asked members to have a look at the paper entitled "Extending the Private Finance Initiative". There is an error in the right-hand column—just to keep members on their toes. I refer to the Jarvis contract with Dorset County Council to build the Sir John Colfox School. I refer to that as being on "a similar scale", but it should read "a smaller scale". Inevitably, the gremlins got in. That project is on a much smaller scale than the other projects to which I refer in that paragraph.

Contracts inevitably require three stages: they must be formed, monitored and enforced. Those three stages are necessary for a proper contract. There is also a spectrum of contracts, at one end of which are hard contracts, in which everything is spelled out clearly and the contract is enforceable in the courts. For that to happen, the contract needs to be verifiable. It must be possible for third parties to see what performance has been undertaken and what performance or activities are required by the contract. If they are not verifiable, enforcement by the courts cannot be secured. Those are the sort of contracts that we have seen under the PFI so far, and that is why they can be enormously costly to set up. If the contract has to spell out in great detail exactly what must be delivered and undertaken under any possible set of circumstances over the next 30 years, it will be an enormous document.

In practice, the documents do not anticipate every possible eventuality—such as September 11—but they do attempt to spell out all outcomes that could reasonably be forecast. That is why PFI contracts usually run to more than 2,000 pages and the contracts that have been created to date have cost more than £1 million in professional advice and fees. They have been extremely costly. One hopes that as experience develops and a greater number of similar types of contract appear, the costs of forming those contracts will come down. Certain agencies, such as the Prison Service, have learned to draw up such contracts quite effectively. I hope that you will have a chance to ask people about that when you visit a PFI hospital.

At the other extreme are soft contracts, which are virtually no more than neighbourly agreements. Incomplete or relational contracts are contracts in which what will happen in all circumstances is not fully spelt out. Such contracts are likely to be renegotiated from time to time. Among the reading is an article by Hart in which that problem is discussed; Kay also talks about that. If everything is not spelt out in advance, it is important that there should be some sort of equality of power between the parties.

If there are outcomes that can be verified and observed by third parties, all well and good—outside assistance to enforce the contract can be obtained. If such assistance cannot be obtained, the contract parties themselves must enforce the contract by any means possible. That might simply be firing—incomplete contracts include contracts of employment, for example. The employees' duties cannot be spelt out in minute detail, nor can the obligation to the employer. From week to week, the boss has to tell the subordinates what jobs are on the agenda. The penalty for not agreeing or not performing will be firing—exclusion from the hierarchy.

With an even softer form of contract, which one of my Edinburgh colleagues refers to as clan relations, the penalty for not behaving as a proper member of the clan is exclusion from the clan—not being acknowledged.

There is a range of contracts, from the hard to the soft. All the contracts have implications—some are costly to form and some are extremely cheap and informal to form. On monitoring, some must be observable by a third party and some might be observable—and even then imperfectly—only by the parties to the contract. On enforcement, those with hard contracts can seek the court's aid. With soft contracts, it is down to the parties to sort it out for themselves.

I will now deal with the reasons for the PFI. I have extracted one chapter from the book that I wrote in 1998, which goes through several bad reasons, several short-term reasons and several good, long-term reasons.

I begin with the long-term reasons. A PFI contract is one where the public sector purchaser does not build the facilities, but finds a private sector supplier who will construct the facilities, put in the staff and deliver a flow of services over the term of the contract, which might be a considerable length of time.

Why is there any benefit in that? First, it requires the public sector purchaser to have a clear idea of what it wants to buy—it forces the purchaser to be clear about the sorts of services it wants to supply and, to some extent, how it wants those services to be supplied. It concentrates the mind.

Under any capital budgeting project for construction of a building, the cost overruns tend to come when the basic terms are renegotiated and variations in the contract appear. That is where the costs really start to build up and it locks in the supplier. Once the PFI project has been agreed, that is that: the facilities will be constructed to deliver in accordance with that agreement. Any renegotiation after that point tends to be costly to all parties, which concentrates the mind.

The supplier that is going to put up the facilities will not be paid anything until the services are delivered and then it will be paid for the services only as and when they are delivered in accordance with the contract. If the supplier supplies late, the payments will come late. There may also be penalty clauses for late delivery—there usually are. That usually means that the construction is speedier than one done under an orthodox capital expenditure scheme.

Incentives are built in for all parties that want the services to be delivered on time and in accordance with the service specifications—in other words, up to quality. The providers are usually divided into a constructor, which is a building firm that puts up the facilities, and a service provider, which is a specialist firm that delivers the services, puts in the staff and ensures that the services are running. The bankers have an enormous incentive to ensure that the project runs, because they will not get their capital back unless the service payments come in on schedule.

The risks are put into the specialist hands of the service provider that has contracted to provide the facility. By the way, much of the literature is not terribly helpful on this point, as it talks about shifting risks from the public sector to the private sector. It is not obvious why that should be a good thing, as the private sector will charge for bearing those risks. The important point to note is that we are shifting risk into the hands of people who have specialist knowledge and who can reduce the risk to make the whole thing work better. Convener, I promise you that I am getting there.

The evidence on whether PFI works is somewhat mixed. I have given members a short extract of the National Audit Office's report, which says, "Yes, it does work in the cases that we have looked at, although there are inevitably teething problems with such a complex scheme." The Treasury task force says, "Yes, it does work." The task force secured a consultancy report from Arthur Andersen, of which an extract is with members' papers. It estimates that about 17 per cent cost savings have been achieved on the projects that were examined.

I must observe that 17 per cent does not sound like an enormous amount. If the first projects are those that are most suitable for PFI, they ought to be the plum cases and one might think that 17 per cent is not an enormous margin. However, we are still learning how to do things.

I have provided a short paper from Unison in the bundle of papers that is before members. Unison says, "We do not like PFI. It is dangerous and leads to an undercutting of quality and the recruitment of staff on inferior terms." I refer members to the papers numbered 5, 6, 7 and 8 in their bundles.

I do not think that PFI is very good for IT projects, as it is difficult to specify IT requirements clearly in advance. A lot of the projects that have not worked have been IT projects. It is interesting to note that on Saturday it was reported that the Inland Revenue has scrapped plans to go in for a PFI project for a major computer facility. Instead, it is going to have a partnership agreement.

The negotiation of PFI tends to be among relatively small numbers of people, who tend to be secretive. It is difficult to make those negotiations more inclusive, open and democratic.

There is a need to protect bankers because, from their point of view, PFI contracts are difficult because they involve their putting up a large amount of cash up front. The only way in which the bankers get paid is if the services are delivered. They cannot control the building of the facility or how the service suppliers carry out the contract, yet they put up the cash up front—ex ante—and have to wait until the project works successfully before they get their money back. The bankers insist that the contract is detailed and precise.

My last slide seems to have disappeared. I have relied on members' indulgence for quite long enough as it is and will leave it to members to ask questions.

I know what I wanted to say on that last slide: I was going to ask whether PFI can work better. I think that there is a case for making PFIs more relational. When I was writing my book in 1998, all the public sector purchasers said that they would like to have more incomplete or more relational contracts but that, unfortunately, the private sector people would not take that risk and wanted everything spelt out. The private sector people said that they wanted more relational contracts and did not want everything spelt out but that, unfortunately, the public sector people would not play ball. The people who insist on the fullest contracts must be the bankers, because they are the most vulnerable.

Remember that there must be some equality of power and that there must be something to gain on both sides in order to have relational contracts that are not fully spelt out. The risk for the private sector is that the public sector purchaser can in the end always opt for Treasury funding and backing, with limitless resources, and appeal to the House of Lords on points of law, whereas it is quite a risk for a private sector provider to undertake to fight the Government if it came to that. From the public sector point of view, the private sector supplier can liquidate, which leaves it in a complete fix.

I have gone through my presentation at a terrific romp; I hope that it has not been too indigestible.

Not at all—thank you very much. That was a very useful way of setting the scene for our questions. I will try to do this in a structured manner and to avoid jumping back and forward to different aspects of the wider question. I invite Richard Simpson to kick off.

Professor Mumford, you said that this system was begun in the early 1980s on the basis of a cost-cutting exercise. What is the current need for a policy initiative such as PFI or PPP?

I think that the current need is to get suitable projects into PFI. A suitable project is one in which the specialised facilities can be outsourced and where the project may be separated out from the core activities of the purchaser, in such a way that there is an identifiable stream of cash flows and some risks can be reduced by the private sector supplier. If a project is suitable, PFIs can provide a very effective way of delivering services.

However, the situation is not the same as it was. There is no longer a public sector borrowing requirement deficit, but perhaps you have a special agenda in Scotland: with the limited powers of taxation here, you might want to view PFI/PPP as a way to circumvent limits on your capacity to spend, in effect through building now and paying later. However, I would not recommend that, unless you have projects that are intrinsically worth putting into a PFI form.

In your article and in your presentation, you indicate a particular factor: assets and services that are unlikely to change enormously over time lend themselves better to a PFI contract. You also suggest that IT projects, which probably account for the most rapidly changing contracts, are probably less suitable. Have I correctly understood the principle of what you are saying? Do you refer only to IT because IT projects have been difficult, or because IT reflects rapidity of change?

It is exactly the latter point: it reflects a rapidity of change. The projects that have been most successful have been those where the technology has remained much the same. The technology of motorway building, for example, is not likely to change much in 30 years. That of prisons might, but that is debatable.

When you were showing us your first slide, you talked about a comparison with private industry outsourcing non-core activities from other organisations. What would you say are the core activities of local authorities, central Government or health boards? I am thinking particularly of health boards, which outsource not only the building of hospitals, but the running of them over the next umpteen years.

I hesitate to dictate what those core activities are and I could not list them, but that is the key issue. I think that the national health service has the greatest difficulty in dealing with PFI, partly for the reasons to which Mr Morgan has alluded and partly because there are so many different authorities, which means that the process of learning how to manage such contracts is spread among a large number of people. For example, somebody in one health authority might know how to do it but then move to another authority. However, I fear that I am getting away from your question.

I do not want to put words into your mouth, Professor, but are you saying that before any organisation in government decides whether to consider using PFI, it should decide what it is that it wants to do and to do best, and then outsource everything else?

That is right. That is a key issue with London Underground. What should the functions of that organisation be and what functions should it buy in? What can it afford to buy in? One problem with IT is that it is extremely difficult to specify one's requirements for 12 months or two years ahead, let alone seven years.

This question is for the CBI representatives. What was wrong with the previous system of public procurement? Even with the developments of the past 10 years, only about 10 per cent of public capital projects come within the PFI remit, so if something was wrong, 90 per cent of those projects are still being done wrongly—if that is the belief. Is that how you see it? If not, how far do you envisage PFI extending? Will it extend to 100 per cent of public capital projects? What figure do you regard as the ultimate target?

We would not put a figure on it in that way. Public procurement has matured enormously over the past 10 years. I would say that the glass is half full, rather than half empty, and it is filling up all the time.

It is only 10 per cent full.

I mean in terms of the sophistication, skill and political commitment of public sector procurement.

I see.

I will pick up on some points from earlier questions that are relevant to this matter. When PFI was launched, there was little political talk about public procurement and little understanding at senior levels in the public sector about good principles of procurement. PFI, almost by fluke, started to identify what is important about good procurement: aiming for best value for money over the life of a project, not the lowest price, and allocating risks where they can best be managed. Good procurement is not about erecting a cheap building and then rolling up high maintenance costs.

PFI is just one form of public-private partnership that tends to put those principles into practice. Across the whole of public procurement there are other ways of putting those principles into practice. The classic PFI deal involves packaging together, when it makes sense to do so, the design, build, financing and operation of, for example, a prison, and rolling all the costs into a single payment to the PFI provider. However, in pure construction projects such as the Egan initiative and the rethinking construction initiative, the same kind of effort is made to put those PFI principles into practice. The risks are allocated where they can be managed and the aim is value for money over the life of the project, instead of the lowest price.

Equally, some local authority deals are very different from compulsory competitive tendering, because in a service-delivery context that may involve less capital, efforts are still made to put the PFI principles into practice.

In a decent partnering arrangement one must get the package of the deal and the model right to implement the PFI principles. One must also have the right political commitment and skills to make the deal work. Different kinds of public-private partnerships are starting to understand how to make those principles deliver better public services.

I do not think that there will ever be a move from 10 per cent of public capital projects coming within the PFI remit to 100 per cent coming within that remit. It will not always be right to package the design, build, financing and operation and have a particular payment mechanism—as in PFI—for every deal. In IT, for example, we are finding different ways of having contracts that put the PFI principles into practice, but a contract might not involve a single payment, as other ways of getting flexibility over the life of the project may be found.

I want to cover two aspects. I heard what you said about there being an upper limit, which you are unable to specify at this time. In some circles, the attitude to procurement is "public bad, private good". Do you generally share that view? Your written submission is pretty gung-ho about the whole approach to PFI, but the first page of your submission says that

"the initiative is still viewed in a generally negative vein by the public at large."

Why is that? Is the message not being conveyed effectively, or is the message defective?

We do not agree with the phrase "private good, public bad". Public-private partnerships are useful, because having a variety of players in the market provides scope for new sources of ideas and healthy competitive pressure on everyone to do better. There is strong evidence of that in the prison sector, for example. The advent of private involvement in the prison sector has caused the core public sector prison service to sharpen up its act on designing and operating prisons. We do not say "private good, public bad"; we say that the market should have more players who are addressing the challenge of improving public services.

You ask why PFI and PPP are not considered good. Procurement has been pretty boring and difficult to explain to the public. It has been hard enough to raise a political debate on the issue. Often, issues become jumbled up. Too often, the questions of who should pay for health services and who should deliver them are confused. Private sector delivery of health care that is funded by the taxpayer is portrayed in the press as privatisation, which creates the fear that the public will have to pay at the point of use. The public debate has been difficult.

It has also taken a while for projects to get off the ground. However, people are very satisfied with the projects that are up and running. A couple of weeks ago, PricewaterhouseCoopers reported on a survey of 27 PFI/PPP projects. It talked to people in the private and public sectors and to service users. The message was overwhelmingly positive. People said that the facilities were much better: pupils have a better learning environment and hospitals have a better healing and caring environment. However, good news stories do not make the press. We have a responsibility to set the record straight and point out what is working. We must also be honest about the lessons that we have learned from initiatives that have not gone well.

Amanda McIntyre said that in many cases the public sector has responded by sharpening up its act. The public sector has decided that if the private sector can do that, the public sector can do the same and something extra. That shows that the process is dynamic.

A threshold of 50 per cent or 70 per cent does not take into account the fact that the process will evolve. The public sector will respond and different forms of PPP will evolve. Projects that would not work as PPPs at present might work in the future. The public sector will learn from the private sector and add its own value so that, in the future, projects that must be PFIs at present will be delivered best by the public sector.

I agree with Michael Mumford that PFI IT projects are particularly difficult, but we must be careful about saying that such projects are unsuitable for PFI and should always be delivered by the public sector. I hesitate to mention the Scottish Qualifications Authority in these surroundings, but the SQA's IT project was a public sector project that proved difficult.

There are many examples of purely private sector IT projects in big companies that have been difficult and had many problems. There are also examples of purely public sector IT projects that have had many problems. It is therefore no surprise that PFI IT projects have also had problems. We must be careful about saying that because one project did not work well, all such projects are unsuitable for PFI. We should work out what the problems were and whether they related to the PFI element.

Are we in danger of following a series of fiscal fads with CCT, PFI, PPP and best value? The law of diminishing returns may operate. You said that the public sector had sharpened up as a consequence of such initiatives. After the cherry picking of the alleged 17 per cent benefit, might benefits be lower for the public sector? Do some of the drivers on the private sector side relate as much to the operation of services as to the design, build and finance? Just how necessary is it to include the operating end of things—which reduces the flexibility in the public sector—to make changes?

I will make two comments in reply to that question. First, I urge you not to forget the problems of cost overruns on publicly acquired projects. It is something that I am sure is ever in your mind.

We have a project under way at the moment. It is never far from our minds.

I could not possibly comment.

That was another fiscal fad.

One of the benefits of PFI is that no payments are made until the service delivery begins. That is a tremendous incentive to get projects up and running—and running effectively—in time. That is a useful feature of those projects.

We are not suggesting that the markets for the supply of all those services will necessarily become very competitive. A useful term, which was developed in the economics literature of the 1980s, is contestable markets, which is the idea that every now and again—perhaps fairly frequently—new contracts will come up and new tenders will be invited. It puts enormous pressure not only on that particular contract but, as Amanda McIntyre said, on all the other contracts that are currently being run. It keeps people on their toes if they know that they do not have a secure job for life.

But there is evidence to suggest that the mechanism of the contest is expensive. In some of the earlier work to assess whether PFI was good value, the costs of the contest were discounted so we did not necessarily arrive at a particularly helpful measure.

I am interested to hear what the CBI has to say about the law of diminishing returns in relation to the adoption of a new approach. I would particularly like to know whether, to attract competition, it is essential that the operation end of many projects has to be outsourced. I was not aware that there were any operators of private prisons in this country. I find it hard to believe that the expertise existed in the private sector.

Underneath the fiscal fads, the public and private sectors are genuinely striving hard to find better ways of working together. That can only be good. It is not the law of diminishing returns; it is the law of increasing returns. We hope that the way in which we do business in 15 or 20 years' time will have matured even from the way in which we do business now.

Improving public services is a huge challenge. The Local Government Association in England and Wales recently did a survey of public perceptions of public services: although 70 per cent of local authority managers said that services had improved over the past four years, only 20 per cent of the public and 23 per cent of businesses agreed. The advantage of involving the private sector is that it provides new sources of ideas.

Recently, I visited some prisons, including private finance prisons and publicly run prisons, and talked to prisoners, prison staff and managers. There is genuine recognition that having different sources of ideas has helped. The Prison Officers Association and public sector prison managers have been thinking again about prison regimes. Prison designs have improved enormously. Prisons are now built with enough work space, so that prisoners do not have to be locked up all day and can do nine-to-five jobs. Private sector involvement in the delivery of public services is producing all manner of service quality improvements. It is encouraging that we are genuinely striving to find better ways of doing business together and to make a much more honest assessment of what has gone wrong in the past.

I will mention some of the work that we have done on the employment agenda. Under compulsory competitive tendering, it was a case of bargain-basement shopping at the expense of service quality and workers' terms and conditions. The CBI, the Trades Union Congress and the public sector never got together to talk about that or admit that there was a problem. Through public-private partnerships, long-term relationships, striving for service quality and continuous improvement, we got together to think about what had been going wrong. Between us, we admitted that there was the problem of bargain-basement shopping at the expense of workers' terms and conditions and we thought about how we could redesign the procurement process to get a win-win—better quality services and a better deal for staff.

I do not say that that work is complete, as there is more to do. However, that is one example of an intractable problem that was causing serious difficulty and pain for many people. We have started to resolve that problem by trying to put those basic—almost motherhood—principles into practice. It is about better value for money, taking a whole-life approach and allocating risks where they can be managed. Underneath the fads, some good things are happening.

In a sense, our inquiry has to look backwards at where PFI came from to see how it might eventually provide a solution to some of the original problems. Why did PFI come about? Was it down to the fact that the old procurement systems were, allegedly, inefficient due to the lack of expertise and knowledge in designing, negotiating and ensuring delivery of contracts?

If we go back to the origins of PFI, it is clear that at that time, public borrowing was quite high and there was a lot of concern about fiscal sustainability and so on. I suspect that the initial driver was to procure assets in a way that would offer better value and lead to reduced public borrowing. Most of the benefits of PFI that Amanda McIntyre talked about have come out of that process. I do not think that those benefits were the original motivation. I suspect that the impulse was the need to find methods of investing in projects in fiscal years in which it was not affordable to do so. A much better approach to public procurement has developed out of that.

The birth of PFI was very painful. The then Government said "deals not rules" and introduced the private finance initiative to get projects through without having the assets on the balance sheet. Almost by fluke, it created an empty space in which the public and private sectors could talk about how they would approach procurement if they had a blank sheet of paper. That is when more challenging thinking became part of public procurement. People recognised that it would make sense to join up design, building, financing and operating. Although that is complicated and has taken a long time to get off the ground, that recognition was the first moment of really radical thinking from practitioners in both sectors.

I agree. The initial agenda was to reduce the public sector borrowing requirement. In fact, an interesting debate has been produced. I have talked only about PFI, which involves a large upfront investment in capital facilities, but there are other forms of partnership between the private and public sectors and they are very promising.

You have been talking about how public sector procurement has improved as a result of many PFI/PPP projects. It is almost an iterative process. You referred to the PricewaterhouseCoopers study, which looked at a number of projects. Did it look only at PPP projects? Has anyone looked at public perceptions of the newer public sector projects compared to the other projects that are delivered through PPPs?

It is encouraging that there is a general debate on improvement of public services. Through initiatives such as beacons, more effort is being made to find out what is going on, even when services are delivered in-house. However, the PwC study looked specifically at the PPP deals.

Is the thrust of what you have all been saying over the past 10 or 15 minutes that the classic PFI—if I can use the word classic about something that has been around only since 1992—has peaked in terms of the proportion of projects that it will deliver, and that such projects are going to exist in a much more mixed market? If so, does that mean that the public sector is learning about improving efficiency from the private sector, and that skills have been transferred?

Definitely, although I hesitate to use the word "peaked". In future, it will often still make sense to combine design, build, finance, operation and remuneration in a single service payment. We hope that there will be an increasing number of models of public-private partnerships that will all try to put the same core principles into practice.

You are saying that head-to-head competition or comparison is going on between PFI and public procurement, in which people must decide whether one is better value than the other. Will not we end up with many models that differ not only in terms of the PFI element in a public works, but in terms of the way in which the public participates in a joint company, for example, in running that company? Why should the public sector be limited to procuring a certain bit of the service on a contract? Why should not a consortium have a public sector partner that is an investor in the project as well?

I have a second question on the not-for-profit trust model, which Alasdair Morgan was going to ask about. Where does that model fit in?

There will be considerably more variety in future. I do not think that the PFI has peaked—there will be many more PFI projects in future, but they will be more devoted to circumstances in which they work very well. There will also be many more of the other sort of model to which you referred, in which public participation exists in different forms. The key feature of PFI is that a specialised facility is built with cash up front, and the incentive is to ensure that it works well. There will be all sorts of other contracts, but if they are not going to be completely defined up front, some balance of power must be kept between the parties.

What about the other methods of getting cash up front? In the 19th century, we used public trusts and public bonds to build much of our public infrastructure. That is now being mooted as a solution for Railtrack, and Manchester airport faces a similar situation. What role is there for that idea? Does not it have advantages?

I agree that there is an enormous role for that idea and for the voluntary sector.

I am sorry if I gave the impression that there is a head-to-head between conventional procurement and PFI.

The public perceive the matter as public good, private bad or private good, public bad. That, however, seems to be a totally false argument.

What we should expect—what we are already seeing—is a number of models of public-private partnership that ought to increase.

On the question of non-profit trusts, it is important to avoid the idea that non-profit-making organisations are a panacea. We would be wary of a flight to the not-for-profit trust model as the solution for every PPP post-Railtrack, because profit is not evil—it makes the economy go round. It is important, in any public-private partnership, that the right model is used to line up all the motivations of all the players to deliver public service improvements. In some cases, there probably is a role for non-profit-making trusts, but that model is not a panacea.

You are right to say that there is a series of motivations that can drive successful public services. One of the areas in which you say PFI would be a big help is in ensuring proper maintenance of capital assets throughout projects' lifetimes, but surely resource accounting and budgeting, which the Government has adopted, will do that anyway. Why is there a need for PFI to deliver that? Just as resource accounting and budgeting will deliver some of the benefits that you say PFI can deliver, other mechanisms will be able to deliver benefits in other areas. Why is it essential that the operating side of the business is in private hands?

There are also situations in which an operation is not entirely in private hands; I am talking about the mixed arrangements that exist in the health service, whereby hotel and cleaning services are often outsourced. In some examples, even nursing services are outsourced and the asset is owned by the public sector. Are there difficulties where different employers are trying to deliver an overall service?

Resource accounting and budgeting is extremely useful in making choices clearer to the public sector. However, it does not guarantee the elimination of political risk to the capital budget. Even more important, under the classic PFI and new forms of PPP, integrating responsibility for maintenance with responsibility for the initial design and build leads to "a stitch in time saves nine"-type benefits. If whoever is responsible for building the asset knows that they must also run it, they will take a whole-life approach to its design and construction.

Would not resource accounting and budgeting deliver the same benefit?

Not necessarily. Resource accounting and budgeting will not always make the public sector client—nor, importantly, the private sector providers—focus on the whole life of a contract. If the design and construction people are motivated to do something that will give value for money over the life of a contract—do not forget that they get paid only if the service is delivered on time, and that the bankers will also be watching what is going on—a value-for-money benefit can be achieved through that joined-up approach.

It is important to manage the interface between publicly delivered services and outsourced services in, for example, a hospital. So far, much effort has been given to improving procurement. The next challenge will be to achieve better contract management and better management of relationships. That is the next set of exam questions that—I am glad to say—the public and private sectors are tackling. I hope that management of the relationship at the interface is high up the list of issues to think about carefully, to ensure that people do not feel that they are in separate teams and that they have no responsibility for keeping an eye on what is going on over the fence.

I would like to ask about resource accounting and budgeting—which you considered in detail earlier this year—and generally accepted accounting practices in the public sector. Is there no evidence that the public sector will begin to think differently? Do you have evidence of the public sector responding by doing things well and in a manner that might, in certain situations, obviate the necessity for PFI projects?

The tide is moving in a good direction. Resource accounting and budgeting, PFI, public-private partnerships and the pressure to improve public services are all stacking up to make the public sector more aware of the importance of whole-life value for money. It is difficult to say that that maturation was the result of resource accounting and budgeting in one place and the result of PFI and PPP in another. It is to be hoped that the clock will not be turned back, but rather that progress will roll forward so that in five or 10 years we will have an even more mature understanding of what is best delivered in-house and what is best delivered externally.

I was not talking about turning the clock back; I was talking about the public sector learning, partly through working in partnership with the private sector.

Absolutely.

Professor Mumford talked about cost savings. If cost savings are not the main driver, the public sector could, having learned the lessons, continue in many cases to deliver much of the procurement.

The advantage of a mixed economy is that everyone gets better at delivery. Some prisons built through PFI were streets ahead of the public sector offering. Some privately managed prisons were also streets ahead of the public sector. However, in a recent competition to deliver privately managed prisons, the contract was given in-house because the public sector option outpaced the private sector.

It is important that the private sector is not regarded as a kind of rescue mission. The private sector must be kept involved for the long term, so that the best minds in the private sector are, with the best minds in the public sector, always thinking about improving public services and trying to strike a healthy balance

The process is dynamic. It is not merely that there are a few lessons to learn, which the private sector teaches for a couple of years until everyone has learned them and we all go back to the old ways; it is about business being a continuous process. The benefit of PFI is that it forces people to think in the long term. Different ways of doing things always emerge, and PFI forces everyone to ask what the best way is.

We support resource accounting and budgeting because it will help to encourage people to take long-term approaches. I agree with Amanda McIntyre: it is not as effective to get a set of departmental ministers to think, "This resource accounting adjustment means that we ought to think about this differently", as it is to force the people who design and build something to think, "I'm going to have to pay my company to maintain this in 20 years' time, so I had better be completely sure that I get it right." Both approaches work together, but resource accounting and budgeting will not obviate the need for the benefits of PFI.

I am sure that committee members have more to read than they care to think about. Nevertheless, I recommend to them "The Contracting Organization" by Simon Domberger, which examines how public sector procurement is practised in a number of different countries, according to several different models. It cites some interesting cases.

I highlight a warning that Domberger gives, which is that to some extent, contracting tends to focus on things that can be spelt out and defined clearly. In the case of prisons, for example, PFI contracts have meant that security of prisoners has been prioritised—there have been very few escapes. Other priorities include feeding people on time, temperature and so on. It tends to be more difficult to identify issues at the soft end, such as in relation to whether prisoners have an education service that will give them some chance of rehabilitation when they get out and allow them to get a job and not just fall straight back into crime. That sort of specification tends to be more difficult to build into a contract. Domberger cites an example from the prison service in America, in which contracting went wrong and the contract reverted to the state.

However, there is no doubt that a learning process is taking place. There is no reason why a private sector contract in the prison service could not include much more counselling, transactional analysis and a variety of therapies that would make prisoners more effective citizens when they get out.

What you say about prisons is interesting. Amanda McIntyre has mentioned prisons many times, and we are going to examine the Kilmarnock prison PFI. The Domberger information will be useful to us. You have been rather sceptical about the success of PFI prisons, but the Confederation of British Industry accepts that PFI in prisons has been a good thing. An Institute for Public Policy Research survey earlier this year found that prisoners wondered whether PFI had been successful. We will have to weigh all that in the balance when we consider the Kilmarnock project in detail. It is interesting that PFI in prisons is emerging as the most commonly cited example in this morning's evidence.

PFI is, as you would imagine, immensely controversial within the prison service, where there are strong feelings for and against it. I have, however, made the point that the prison service has learned very well how to manage its contracts.

We have moved to whole-life management—that is the stage that we are at this morning. That will produce certain tensions, the primary tension being between getting specifications correct at the beginning of a contract and building in mechanisms for alteration. I do not mean changes just in technology, but in the broader learning process. In setting up tremendously long-term projects, it is vital that the original specification—if it is to be a hard specification—is all encompassing, which is perhaps too bureaucratic. Is there expertise in the public sector that can handle setting up contracts at the very beginning? Is there enough expertise around for people to be able to plan ahead and negotiate that flexibility?

Not completely. The body of expertise is still growing, but there are capacity problems and also some issues that we have not quite cracked. As I said, much thought has gone into procurement and the next series of issues will concern management of relationships. There are some innovative ideas concerning the best means of managing relationships and achieving overarching frameworks, and a process for agreeing how changes might be handled over time.

Highways agencies provide a good example. A team has learned how to set up such contracts, and some motorway contracts are very effective. The motorways, bridges and facilities are being built, the suppliers are being paid according to vehicle miles, the technology is fairly clearly defined and motorways are being managed very effectively in most cases.

If there is a vacuum now, what needs to be done to step up the process?

Civil servants who handle PPPs must be valued. There must be clear career progression for them in that market. The attitude must not be to see them merely as boring old procurement people who can be pushed into a cupboard somewhere and expected to do something very nice for us. We must not say to such people, "You've done so well in that boring old procurement job. Here's something really interesting. Why not take a bill through the Scottish Parliament? You'll get real brownie points for that." We need to get away from that mentality. If somebody is delivering PPPs and improving public services, that person should be valued as part of the senior team and rewarded for that. The saddest thing is when somebody builds up expertise and then teams are fragmented and dissipated.

That is because of movement within the civil service.

Exactly. The committee's inquiry is extremely useful. It is important that politicians continue to take an interest in the matter and—as members are doing today—that they bother to get to grips with the technical detail of how it all ticks.

Has movement of people just as they have begun to grasp their subject been a big problem?

Yes.

It has been a problem especially in the health service.

Expertise in the health service is fractured anyway. We need a central team in Scotland; there used to be an advisory group. Expertise must be built up and people should not be moved to another department as soon as they gain expertise in one. As part of the loosening up, should not people be more easily transferable in and out of the public sector? People in the private sector can get into the public sector and be on that side of what we hope is a diminishing fence; they can work more in partnership.

Absolutely. I could not agree more.

We have not got to the end of some of the current contracts and there can be peculiar problems towards the end of contracts—especially in valuing facilities and either transferring them to the public sector or renewing contracts. There will be difficult issues of that kind. At the moment, the contracts are very large—they are held on CD-ROMs—and the only people who know them inside and out are the bankers. If matters were to come to a nasty legal conflict, it is likely that the people who know the contracts would win the legal arguments in court.

We have referred several times to value for money, which is the criterion against which PFI and PPP projects are assessed. What are the different ways in which to determine value in such contracts and projects?

We have also talked about the life of projects and about how maintenance costs are built into them. Does that form a key part of achieving value for money? We also need to ask who gets that value for money; is it the consumer or the Government? What are the issues?

You talked about the assessment of soft parameters, within the Scottish Prison Service. Such parameters are much more difficult to measure and assess effectively, so how do we proceed with them?

That is a key question for a public sector client. The definition of value changes over time. I suspect that 20 years ago, quality of public services and patients' experiences of it were not valued as much as they are in today's political climate. An Executive or Government that was deciding on value for money in the current climate would want to put a high value on that sort of quality-of-service provision. That must be covered in contracts. We have discussed the difficulties in that, but—as Michael Mumford said—there is a learning process.

There are examples of reoffence rates being included as incentives in prison contracts, which is a neat way of saying that the public sector values the integration of prisoners back into society. Such integration is difficult to manage, but attempts might be made to tie contractors into that. It is partly for politicians to decide what they value and it is then for the private sector to work with the public sector on how best to deliver that.

We need to widen the analysis of value for money and to spend more time thinking about how we value quality, not just cost. We need to think about value for money over time.

Let us go with the contestability argument. If Matthew Mumford and I are bidding for something, there might not be very much between our bids, but our bids today are probably much better than they would have been five years ago, simply because of the healthy competitive environment that we have created. We need also to factor that into the analysis.

I have no doubt that that is essential. As I said, there is always a danger of focusing on what is easily measurable. All the press reports on the PFI hospitals have despaired about the cut in the number of beds. That is easily measurable. Laparoscopic techniques and the increase in day surgery mean that probably fewer beds are needed for conventional patterns of service; more beds are, however, needed for geriatric cases and for people who cannot find care homes.

The CBI's paper refers to the value for money that was achieved in the A74 project. That view was not shared by the National Audit Office nor by the Scottish Parliament Audit Committee, which suggested that

"there was at best only marginal financial gain from selecting the privately financed road over a publicly funded alternative".

Are you aware of the CBI report, and of the detailed concerns that were expressed by the NAO and the Audit Committee about whether we were getting value for money? In particular, they had concerns that public sector comparative figures were £8 million out. There were arithmetic problems in that the costs of the procurement exercise were not included, which knocked another chunk off the value. Despite early delivery, an incentive was paid that appeared to be out of all proportion to the benefit that was gained. That almost certainly meant that the sums added up to a negative figure. The detailed approach, which involved trying to keep more people in the bidding process for as long as possible, meant that there were significant additional costs in the project.

I am not an expert on the detail of the A74 project. As Michael Mumford has said on a number of occasions, there is a learning process. It is difficult to work out the best way to manage a contract. Private sector companies often say that bidding is incredibly expensive; they can spend about £500,000 preparing a bid for a major project that they end up not getting. There is a lot of learning to be done on both sides, but the question is whether the concerns show that the principle of PFI is flawed—I argue that they do not—or whether they show simply that the process can be better managed.

Professor Mumford said that one of the weaknesses of PFI was the excess of bureaucracy. Will that decrease as we get more sophisticated partnerships and soft relationships? Will that reduce the size of the contract documents? You said that some of them ran to 2,000 pages.

The early PFI documents were often 2,000 pages long and came on CD-ROM—they were searchable on a computer.

Two forces are working in opposite directions. There will be more variety as there will be differing sorts of contracts but, within those varieties, there will be more standardised contracts. When people have used a couple of contracts, they will know how to do it again. As somebody would go to a solicitor to get a formal will, it will be possible to get a standard contract for a short-term IT or schools management project.

That will help a lot and will, I presume, reduce transaction costs.

I want to move on to contracts. Professor Mumford says in his article that PFI's weakness is its excessive bureaucracy. In your opening remarks you said that bankers were most keen on using detailed contracts because they are the most vulnerable people in the PFI process. I do not see how it is possible to get away from having detailed contracts, but how can excessive bureaucracy be dealt with when the bankers, who are at the heart of the process, insist on that level of bureaucracy?

The classic sort of PFI, which involves a large capital expenditure up front that can be recovered only from the payments for the services that will come in over the next 20 or 30 years, will always need to have a complex and full contract. If we are going to use more relational contracts, bankers will need other forms of security. That will not be difficult if there are alternative uses for the assets, because they can be used as security for the bank loans. It is possible to take out a mortgage on a row of buildings, but if there is only one use and one possible purchaser for those buildings, the banker cannot use them as security.

If we want to ensure that a good quality public service is delivered, should that be written into the contract? Obviously, certain minimum standards will have to be established, but how would it be possible to get a high-quality service without including considerable detail in the contract?

It is possible to observe performance and tidy it up. Within a firm at the moment, there is a hierarchy. People observe how work is being done and refine the way in which it is done by telling people or using an appraisal interview to point out areas that could be improved. That is a relational contract—similar arrangements are used all the time in people's jobs.

In my paper, I point out that there are bureaucratic problems with the classic PFI, which involves a big slug of money coming in up front. However, the problems are lessening as we become more familiar with the system.

Amanda McIntyre says in her paper that PFI projects offer the opportunity for more flexibility in decision making. How can that be written in contractually?

We might think in terms of output rather than input. That would give the private sector provider more flexibility and scope for innovating on delivery. Partnering approaches are coming more into practice. A contract underpins the core criteria but, typically, the public and private sectors form informal partnership boards as a mechanism at a senior level for the client and the service-providing partner to talk about the need for changes over the life of a contract. These are fairly early days—most deals are new, but much greater effort is being made to embed a partnering approach at all levels of organisations.

Did you say that those partnership boards are informal?

Yes.

Problems arose with some CCT contracts because of clients' lack of experience of managing variations. Many contractors ripped clients off. Would not a more relational style of contract carry the significant danger of that happening again, especially if a single contractor deals with the continuing contract and variations in that contract? All the power in that relationship would lie with the contractor, rather than the client.

There is a world of difference between CCT, which involved pricing low and coping through variation orders, and robust, mature and modern partnership arrangements that involve much planning up front. Much more of that is happening before a contract is signed, and that is followed by sensible arrangements to cope with changing requirements over the life of a deal.

The public sector client has more power than you might think, because the private sector's reputation rides on delivering a good service. As the public sector becomes more mature as a client and more sophisticated at choosing between bidders, a good track record will stand for much. Service providers are in the market for the long term.

I understand how what you describe might be valuable in a relatively short-term contract, because it allows a change to be made. However, if a client has chosen the wrong contractor for a 30-year contract and is dealing with a fairly open contract, all the power will lie with the contractor. You are trying to sell the system to me on the basis of goodwill. The client needs a little more power to arrive at a fair arrangement. I suspect that with a single contractor, the balance of power on variations is unlikely to be to the public's benefit.

It is only when a very large investment is made in specific facilities that a serious problem exists. Otherwise, there is no reason for not having fairly short contracts, for which further tendering exercises can be undertaken. The power of the purchaser is considerably greater in such circumstances.

It is interesting that suppliers and purchasers have said that some of the early PFI contracts are too long. They are not contestable. The parties are locked in and there is no way of changing those contracts. Relational contracts would be unlikely to continue for as long as 30 years, unless an enormous commitment to specific facilities had been made. That would require a pretty complete contract to protect the parties.

No witness has mentioned performance indicators. What role do they have in the loose relationships that you describe?

The situation with loose relationships is interesting, because they allow performance indicators to be renegotiated. In a relational contract, it is up to the parties to redesign performance indicators from time to time. The only way of ensuring that the parties play the game and are genuinely in partnership is to make both of them see that they have much to gain from the relationship's continuing.

Performance indicators allow the public who receive the service to assess whether they are receiving what they were sold in the beginning.

Performance indicators are important. We have not mentioned them because it is hard to lay down general principles. The only general principle is that the clearer the issue is, the more precise the performance indicator should be. As I said, the danger is that people concentrate on performance indicators and lose sight of some softer issues. We tend to concentrate on what can be measured.

Professor Mumford said that bankers would be more likely to be happy with relational contracts if they had a marketable asset at the end of the contract. If an asset will not be marketable, is the corollary of that the fact that bankers will charge a fair bit over base to lend the cash? In that case, would not more traditional methods of public funding be more viable or suitable?

That is precisely why PFI was introduced. It was developed for highly specific assets that had only one use, such as a motorway junction that might have virtually no use other than to join the M4 and the M6. That is where PFI began. Such projects required full and detailed contracts that specified exactly how everything would work. In the initial years, the interest rates on such projects tended to be high, but once contracts were up and running and became established public sector projects, interest rates dropped fairly dramatically.

People have made the plausible argument that if a motorway junction must be maintained for 30 years, a good job should be made of building it. We have never reached year 25 of a project, but at that point, what is the contractor's incentive to maintain the facility well, rather than to patch it? The contractor might not be interested in winning the contract again or might have so many competitors that the chance of winning it again is not high.

That is a tricky problem. I have tried to obtain operational research to resolve that question, and I have not yet found the answer. Many contracts encourage the private sector supplier to run the facility down in the last years of the contract, so that the facility will collapse on the last day of the contract. It is difficult to avoid that. The key to the uncertainty is the question whether the client will sign up with the same provider for the next period or switch to another provider.

We have talked quite a lot about quality. How do we ensure that quality is maintained throughout a contract? People at the coalface—or in the hospital bed—demand quality service. What systems ensure that efficiency gains do not take priority over service quality?

That is ensured in PFIs by spelling everything out in the contracts. The service specification is precise and the service provider is not paid if the service is not delivered or is not of the required quality.

Relational contracts are more difficult, because not everything is fully spelled out. Renegotiation must include specification of service quality, which must be observed.

I have a particular interest in IT. It is inevitable that that will involve relational contracts. It will be difficult to say what quality of service means, because the definition now may not be the definition even in five years' time, never mind a bit further down the road. Who should regulate PFIs and PPPs? Who has primary responsibility for that?

I see that that question is coming my way. That matter is down to politicians. Whether to pursue that is a political decision.

The relationship between the public and private sectors in a PPP or a PFI is one of client and service provider. It is important that the public sector is an excellent client. The new PPP contracts, not just the PFI contracts, have strong performance management regimes—much stronger than ever before. As Professor Mumford said, if the service is not delivered to the required standard, the contractor does not get paid. Complicated performance indicators run through PFI and PPP contracts.

It is important to unpack what we are talking about when we talk about relational contracts. The choice is not between PFI and a relational contract; it is between PFI and other forms of public-private partnership. The latter may have a contract, but be administered in a different way—there may not necessarily be a unitary payment for every prison cell that is built, for example, or for every car using the road that is built. Such contracts will be just as binding. Professor Mumford spoke about looser relationships; they tend to be more suitable for smaller scale partnering arrangements in which there may not be so much business and for which one particular provider may be the provider of choice—the first place the client would think of going for the service.

For some long-term contracts—especially in the health service—new quality standards will come in as a consequence of future events. An example would be a hospital-acquired infection; it could be that a contract is entered into that does not take such an event into account, but then the event happens and is a major problem. How do we cope with that? I am thinking more about the operational side than the design, build and finance side. There will be a 30-year contract, but we may wish to have a much shorter contract for cleaning. Such a contract will be totally different from the contract for the maintenance of the building and equipment. Do we need to be a little more sophisticated in the way that we divide contracts up? New standards of quality will arise during the lifetime of a contract, so how can we ensure value for money for the public when we have a long-term contract that may not be flexible and for which the power remains with the contractor rather than the client?

Before contracts are signed, there is agreement over which party carries which risk. There is agreement, for example, over whether the contractor carries the risk of any extra costs or services that new environmental legislation may impose. There is discussion of how various risks will be carried and it is all written into the contract. A new set of requirements may trigger a discussion over a new service or a price change, but things will have been agreed in principle at the outset. That means that everyone will have already agreed on a way of dealing with any new event. If such an event happens, people will go back to the contract and say, "Okay, this is what we agreed. Does it still make sense? Yes, it does." That kind of arrangement gives the client some power.

That approach seems sensible. Are you aware of that approach having been adopted in the significant number of PFI contracts for hospitals in Scotland? In some of those cases, the build is past; in others, it continues.

The PFI standard contract guidance sets out ideas on how to handle all such issues, so that each NHS trust does not have to think things out from scratch. The current set of standards is, quite rightly, under review, to see whether we have to update it. However, I think that most hospitals have been following the standard approach.

Mr Adam asks what happens when something unforeseen crops up. Contracts try to foresee the major things. Interest rate changes, inflation and all that sort of thing are built in. I have tried to give the impression that under the complete contracts everything is foreseen; of course, in the real world, not everything can be foreseen. There is a tradition of renegotiating from time to time. If there is good news, the benefit should be shared between the parties and not kept by one party or the other. If there is bad news, it should be borne by both parties and not just one or the other. There has to be some element of renegotiation, although that can be a bit costly in big contracts. However, it is often done. At Colfox School, all sorts of things happened that had not been anticipated when the school was set up and built. For example, kids would get caught in a shower of rain and the caretaker would open the school an hour earlier than usual. The contract does not make any reference to that happening, but there has to be an element of give and take.

The answer to a recent parliamentary question of mine told me that there has been a dramatic roll-out of PFI projects. That raises the question of whether there is genuine competition out there. Are there enough players to cope with demand? Do the witnesses feel that there is enough competition and, if not, how can it be generated?

This is a serious issue. If you think you have problems in Scotland, think for a moment about the problems of the Ministry of Defence with an enormous project such as the procurement of fighter aircraft, for which there are only two or three plausible providers worldwide. It is a problem. Refining that problem leads to serious questions that we have not really resolved: how many people should be invited to bid; to what extent should a preferred bidder be encouraged; should that be permitted or does having a preferred bidder make a mockery of the whole process? If there is a preferred bidder, other bidders are liable to say, "This is not really an open competition because we all know that the contract will go to Siemens," or whomever it may be.

It comes down to specificity. If you are talking about very specific inputs, there may be a limited number of people who are capable of doing the work, but if you are talking about something fairly general, you will have no real difficulty in sourcing the work. In the latter case, questions arise of how many bidders will be sought and how long discussions will continue before a firm contract is arrived at.

I presume that such issues would still exist under the old procurement system.

Yes, exactly.

There is competition from UK firms and international firms. The market is growing globally. Lots of countries in Europe and further afield are considering PPP approaches. That can only be good. However, it is sensible for the Government always to behave as a challenging client and to avoid the situation where PPP deals are like buses—you wait for one and four come along at once—because that, clearly, does not lead to best value for money for the public sector.

I return to my point about maintaining a deal flow. If the Government wants the private sector to gear itself to the needs of public services, the Government has to signal to the private sector that there will be a long-term market so that people can put their research and development efforts into public services as opposed to anything else.

Will there be an unbundling of contracts or an amalgamation of contracts into major exercises?

A combination of both. In health, education and local government, it can make great sense to have more collective procurement for common systems and services. On the other hand, as we get cleverer at doing PPP and PFI deals, and as transaction costs fall, as I hope they will, the tendering costs of a partnering approach may not be so prohibitive. The threshold at which it becomes feasible to do a PPP deal may start to fall.

I want to discuss staffing issues when there is a change from public to private procurement; I want to explore the social context. I am interested in the comments on staffing in the submission from CBI Scotland, which mentions

"ensuring that the goal really is best value for money as opposed to lowest price"

and ensuring that the Transfer of Undertakings (Protection of Employment) Regulations "are applied consistently". In my experience, the issue is not the consistent application of the TUPE regulations, but what happens after that. There is not much difficulty in an individual or a group of individuals transferring across and maintaining conditions. The problem is how to improve the conditions after that and, more importantly, how to deal with the damage to morale once new staff are employed. It must be damaging to morale to sit next to a person who does the same job yet has different conditions simply because they were employed a month after the transfer took place.

How do you think that staff morale—which you mention in a positive vein in your submission—can be maintained? Are you in favour of going as far as the Secretary of State for Trade and Industry who, when speaking to the TUC conference earlier this year—in September, I think—mentioned the possibility of extending the TUPE regulations to secondary transfers, to cover not just pension provision, but general conditions? Should that be backed up by legislation?

We favour the consistent application of TUPE. I would contract to Matthew Farrow and the staff would move across. If, five years later, Mr Farrow lost the contract to Professor Mumford, the staff would move across again. In the past, there have been problems with staff moving across on the initial outsourcing and not subsequently. That is perhaps an example of the sort of opportunism that Professor Mumford talked about. The Cabinet Office guidance—which we were instrumental in putting in place—makes it clear that that should not happen. Staff should transfer consistently.

In the past, the bargain-basement shopping experience has been the biggest problem. The way forward is better procurement—ensuring that the public sector aims for a quality service and is prepared and able to pay a fair price for that service. The public sector client must choose the contractor in a way that takes proper account of the relevant work force issues. As well as doing business with somebody who can manage a work force well, the client must ensure quality and achieve a win-win situation through better procurement.

On more than one occasion, you drew a line under compulsory competitive tendering, which you said was purely for cost saving. You looked at private finance initiatives in a different way and what you have just said expands on that. I return to how you, as an employers' organisation, would cope with the situation that I cited earlier, in which two people who sit side by side have different conditions for doing the same job. Rather than just good will, is not it necessary to have some form of legislation to ensure that that does not happen?

I do not think that employment legislation is the way forward on that. Large service delivery companies might have a range of terms and conditions in place and it is their managerial responsibility to organise that situation. However, we need to get away from undue pressure on contractors to cut corners through the lowest price approach.

We having been thinking through the handling of the procurement process with the unions and the public sector. When considering the award of a contract, what sorts of employment issues should the client talk about with the contractor at that stage? How can the client ensure that the contractor delivers a quality service and is able to recruit, retain and motivate staff of a high enough quality to deliver the service that is demanded? The answer must lie in better procurement, rather than more employment legislation. The procurement issue has been the root of the problem in the past. Therefore, better procurement is the solution for the future.

Professor Mumford, in the section of your presentation entitled "Does the PFI work?" you said that Unison was opposed to the concept of PFI. How would what you have just heard—for example, Amanda McIntyre on behalf of the CBI—sit with dealing with the sort of staffing problems with PFI projects that you are aware of Unison raising?

I share those worries to some degree. I am not even sure that, if legislation existed, it would be easy legislation to police. The sorts of things that concern me are situations such as inputs being sourced from, for instance, sweatshops in Indonesia rather than being made in this country or other parts of Europe, which are properly regulated. That is a worry.

Amada McIntyre is right that it will be difficult to guarantee maintenance of service up to service specification if we use the cheapest possible sources of labour. That tends to mean that items that are specified in performance indicators are met and the soft things out at the end all fall off. That is referred to as service shadowing.

I would be concerned if staff conditions, for example, were to be described as the soft end. Staff conditions are a central part of a service specification because the staff ultimately deliver the service. If they are not motivated, that will be a severe problem.

I too have concerns about that.

Is it possible to state in the procurement process that, if a company's staff turnover or sickness level is above a certain level—in other words, if there is evidence of poor management, which is what those two factors and many others would be—the company is failing to meet its contract?

I will give another illustration. Effectively, the previous Administration privatised care of the elderly. The result is that we are now talking about raising the standards through, for example, the national care standards committee, and yet the funding of that sector is poor. The complaint of independent providers is that the minimum wage is giving them problems. It should not be giving anyone a problem, but it clearly is. How do we ensure that, in a sector such as health care, the staff are valued and have terms and conditions that ensure that they provide what used to be called the public sector ethos type of approach?

Your example illustrates my point perfectly. Consider what local authorities have been paying the independent sector to deliver residential and domiciliary care. I only know about England, but there it has not been enough to cover the cost of delivering the service. We often find that local authorities pay the independent sector about half of what they pay their in-house service providers to deliver a comparable service.

First, we must have a real commitment to value for money, not the lowest price. That raises a set of questions. If, say, I am thinking about giving you a contract, Dr Simpson, what sorts of questions do I ask you to ensure that I am confident that you will deliver a service and treat your staff well? Good treatment of staff is in my interest because I want a good service, but also because I have a moral and social obligation to ensure that you are treating your staff well.

We have been having constructive discussions on that matter with the unions—quietly behind the scenes. When is it appropriate for me to ask in-depth questions about health and safety and about training plans? When I hand the work over, will I be honest about the current service level and the service level that I want? Am I honest about that or do I pretend that the current service level is different, so that you budget to deliver a certain amount of service when, in reality, when you inherit the work, you will have more work to do with your current budget. Issues such as those must be tackled honestly to get the right result.

You ask in your submission why PPP and PFI is regarded so badly by many of the public, particularly those with a public sector background. The answer is that the exercise is about degrading and devaluing staff—a bargain basement approach, as you said. It has been all about that. That is why the unions are so against PPP and PFI. They do not regard staff as having been protected. TUPE has been wholly inadequate.

There are, however, numerous examples of staff moving across to the private sector and having a far better time.

They have had better working conditions, for example.

Will you give us some examples of such instances in writing at some point?

I noticed that your submission referred to such instances. It would be helpful if you could send the committee some information on that.

A reader of Private Eye might be tempted to ask whether people involved in the security of Scottish schools might have contracts that ban the employment of staff who have a criminal record. That is a dangerous condition to insert into a contract too crudely.

There is an interesting discussion of prison services in the Domberger article at page 166 onwards. The article refers to a code of conduct that was drawn up by the American Correctional Association and that lists criteria such as the 38 mandatory standards for accreditation. Applicants must also meet 90 per cent of the non-mandatory standards. The list of things that a good prison service must have is quite long. Thereafter, it is up to the public sector purchaser to insist on an ACA-accredited service provider. That sort of approach is possible.

It was helpful of you to point that out—we will consider the article in detail.

I thank the witnesses for the full answers that they have given us over the past two hours and for their invaluable contribution to our inquiry. They have given us a useful start and we hope that they will track our progress over the next few months. As they may know, later on in our inquiry, our work will include visits to PFI projects. We will ensure that they receive a copy of our report at the end of the process.