Public Audit and Post-legislative Scrutiny Committee

Meeting date: Thursday, February 20, 2020

Official Report 556KB pdf

---

Section 23 Report

---

“Privately financed infrastructure investment: The Non-Profit Distributing (NPD) and hub models”

Item 2 is consideration of a section 23 report. I welcome our witnesses, who are all from Audit Scotland. Caroline Gardner is the Auditor General for Scotland, Mark Taylor is assistant director of audit services, Graeme Greenhill is a senior manager, and Martin McLauchlan is audit manager for performance audit and best value.

I understand that the Auditor General has an opening statement.

Thank you, convener. This is a complex subject, but I will keep my introduction brief.

The Scottish Government invests in infrastructure assets including schools, roads and hospitals to support delivery of high-quality public services and economic growth. Privately financed contracts allow the Government to make additional investment beyond that which is supported by capital funding in the block grant. The Government recognises that such contracts are more expensive than traditional public financing, and it accepts the extra costs to enable the additional investment to go ahead. However, it is important that the additional costs are understood.

Most investment is still financed through traditional means, but the amount of privately financed investment is substantial. Since 2005, the Scottish Government, working with councils and other public bodies, has used NPD and hub contracts to build assets worth £3.3 billion. That is on top of the £5.7 billion-worth of assets that were built under earlier private finance initiative contracts. Payments on those contracts totalled £13.1 billion up to the end of 2018-19, and the public sector is committed to further payments of £27 billion through to 2048.

PFI, NPD and hub contracts have been used across the public sector. The Scottish Government decides when they will be used by determining its capital allocations and offering to meet all, or a substantial proportion of, the on-going payments. It is not currently clear enough how the Government decides which projects should be privately financed and which will use traditional public financing, which means that it is difficult to assess whether the combination of projects and use of private finance is delivering value for money overall.

NPD and hub contracts are no longer used for new projects, and the Government is introducing two new funding mechanisms in their place. The first of those—the mutual investment model—aims to allow continued use of private finance to provide additional investment following changes to the national accounts rules.

The second model—the learning estate investment programme—will remove the need to directly access private finance for schools projects by linking on-going Scottish Government contributions to the achievement of specific outcomes.

The models are being developed and introduced alongside the recent report by the Infrastructure Commission for Scotland, with a new infrastructure investment plan being due for publication later this year. The Scottish Government needs, alongside those developments, to develop its reporting to Parliament and to the public to show how use of private finance is contributing to its wider policy aims.

As always, the team and I will do our best to answer the committee’s questions.

Thank you very much.

I will start with a question for technical clarification. Obviously, the Government cannot use capital funding to fund revenue expenditure, but it can use revenue funding to fund capital programmes. Maybe I am missing something, but I do not see that mentioned in the report. Such funding has been a source of capital funding in the past. Have you looked at that, even though it is not mentioned in the report? How does it compare with the other ways of funding capital programmes?

The report looks specifically at private financing for infrastructure investment in the context of the Government’s wider capital powers. I do not think that we have a figure to hand for how much is being financed from the revenue budget. It will be a small amount, but we can come back to the committee on that, if that would be helpful.

That would be helpful for comparison. The report very usefully compares the costs of funding through PFI, NPD and other such models with the costs of funding through the Public Works Loan Board. It would be interesting to see the impact of the conversion from revenue to capital and whether there is more scope to do that in the future.

The costs include the opportunity cost of public services not being financed in the normal way, which is an area that has been under pressure for the past decade or so. We will look to see whether a figure is readily available.

I draw your attention to paragraph 48 on page 24 of the report, which is interesting. Your analysis shows that, compared with traditional funding through the Public Works Loan Board, funding schemes such as PFI and NPD lead to additional costs of up to 5.7 percentage points higher. Can you quantify the additional costs of using such types of funding? Over a 30-year period, for example, up to 5.7 percentage points is a lot of money, which reeks of people ripping off the taxpayer through such schemes. I want to explore the issue, because it definitely looks like a rip-off.

I will bring in the team in a moment. As I said in my opening remarks, the Government recognises that using private finance incurs additional costs. The Government accepts those costs because of the additional investment that it enables, the economic growth that it believes is generated and the ability to support public services.

I think that we all know and accept that using private finance involves additional cost, but those additional costs are up to 5.7 percentage points per year. There is a difference between accepting additional costs and being ripped off.

That is why we make the point that the Government should be clear about how it decides which projects are being financed privately and about the wider economic benefits that they are generating.

We cannot quantify the additional costs here and now, but I am sure that we could go back to the office and do some calculations for you.

That would be very helpful.

I know that there is a range, but we know from other research that in many projects the additional costs are nearer 5.7 percentage points than 1.3 percentage points. If you did a breakdown of the PFI, NPD and hub projects, would the range be higher in the PFI projects than the NPD projects?

We have not done such analysis. We looked at NPD and hub projects. In paragraph 18, we say that “caution must be exercised” in undertaking such analysis, because the margins are driven not only by the market rate at the time and the opportunity cost to the investors, but by the services that are included in the contract. Some of the surrounding headline figures of the ratios of the capital value to the unitary charge aggregate would be influenced by removal of certain services. Our point is that paragraph 18 should be noted.

We also make the point that there is a difference in timing, in that most PFI projects happened in the run-up to the financial crash in 2010, and most of the NPD projects happened after that. There was a difference in risk appetite and interest rates after 2010, which we have to take into account.

After the crash, however, interest rates crashed. The top of the range is a rate of 5.7 percentage points. If that has been the rate for privately financed projects since the crash, that is even more of a rip-off—this would be true even if we took the average of the range—because at times during the past 10 years we have had negative interest rates. Even now, the Bank of England’s base rate is only 0.75 per cent.

You are absolutely right about the impact of the crash on interest rates. I meant just that it is another factor that makes it harder to compare PFI projects with NPD projects. They happened in different time periods, so we are not comparing like with like.

Elsewhere in the report, you rightly make the point that there is a “lack of transparency” in terms of senior and junior debt. Am I right in saying that the private sector is making a lot of money through the excessive interest rates that are being paid, and from selling debt once projects are up and running? The latter has happened in many cases, of which Hairmyres hospital is a good example.

Again, I will bring in the team, in a moment, to respond.

You are right to say that there are two elements to that. The first element is the interest rates that are being paid, across the life of the contract, to the special purpose vehicle and to the investors who are part of that. The second element is gains that are made over and above returns. The development of NPD was intended to cap those returns; it has done so by removing the opportunity to sell on equity in that way. However, as we say in the report, there is a market for the secondary debt and there is still the possibility for returns.

Mark Taylor can give you a bit more detail on that.

A fundamental part of the NPD process is that the amount of interest is built into the deals—that is set and agreed at the start. In the report, we say that in order to understand the extent to which that interest has been adjusted to reflect the fact that companies cannot take dividends out of those deals any more, we need to look at the secondary market and at how trades are being made. We need to glean what we can from that in order to understand the extent of the profit that was built into the initial pricing. Many factors drive that initial pricing. In the report, we set out some of the elements that keep those numbers down, including levels of competition and suchlike.

The problem that we have—it is probably a problem that you have—is that many of the details are secret and are hidden behind the rule of commercial confidentiality. Should all that information be in the public domain, given that we are talking about taxpayers’ money?

One of the fundamental recommendations in the report is the need for greater transparency about when and how privately financed vehicles are used. I am not sure that we agree that all the details should be in the public domain—certainly, at the outset, there is an element of commercial confidentiality to consider—but there is a need for greater transparency and for the Government and/or the Scottish Futures Trust to monitor the secondary market that Mark Taylor mentioned. At the moment, they do not do that, but it would provide useful information about exactly the levels of return to which you refer.

I would have thought that the basic two pieces of information that we should know and that should be in the public domain are who owns the senior and junior debt, and how much profit they are making from a contract. I think that, as taxpayers, we are entitled to know that.

The market in secondary debt is a useful indicator of that. It would be useful to monitor and publish more about that.

Am I correct in saying that, under all NPD and hub projects—unlike with PFI—the assets remain with the public sector at the end of the period?

Yes.

That is correct.

Obviously, we are in the process of setting up the Scottish national investment bank, which, I think, had universal support across the Parliament. That will primarily fund private sector projects; it is not to do with providing public services. Should there be a role for the bank in funding infrastructure projects, for example?

You will not be surprised to hear me say again that that is a policy question for the Government, rather than one for me, as the Auditor General, given my remit. However, it is worth the Government’s while to think hard about the different ways in which it invests in infrastructure and how it aims to promote economic growth, given that that is such an important rationale for use of such contracts. Those contracts need to be aligned and to make best use of expertise, and to join up with the outcomes that the Government is trying to achieve, and the Government needs to evaluate how well that is being done.

There is an issue to do with the 5 per cent rule, which is set by the Scottish Government, not by the Treasury. According to you, less than 4 per cent is being used. Therefore, about one fifth of spending capacity is not being used, which, in the current economic climate, is rather discouraging. In my view, we should be using it to near maximum.

On the ability of councils and the Scottish Government to borrow from the Public Works Loan Board, the Scottish Government is limited to £3 billion, and councils can borrow proportionally more that the Scottish Government. What share of its capacity to borrow is the Government using? What about councils?

09:15  

Mark Taylor will keep me straight on this, but I think that the Scottish Government is using its full capacity to borrow and intends to do so under the medium-term financial strategy. You are right that there is some headroom—if I can use that word—within the 5 per cent revenue limit that the Government has set. That limit is useful.

The national infrastructure mission intends to raise the level of capital investment by, I think, £1.6 billion in order to increase the economic stimulus that Alex Neil is talking about. Councils operate under quite a different regime—the prudential framework—which enables them to borrow according to the amount that they think they can afford to service and repay. It is worth thinking about that in the round.

As we say in the report, being clear about which financing mechanisms are being used for which projects under the national infrastructure programme would be a step forward in relation to transparency and demonstrating value for money.

If we were nearer to capacity for Public Works Loan Board funding, the vast bulk of the £1.6 billion would need to be funded through other means, which are the most expensive.

That is right, and it is one of the reasons why the Government is continuing to develop the mutual investment model.

I presume that we should also ask the United Kingdom Government to allow us to borrow more through the Public Works Loan Board, because we would then get better value for money.

Clearly, the fiscal framework is one of the constraints on the amount of money that the Government can spend overall.

Absolutely. It is a nonsense.

I entirely agree with Alex Neil’s line of questioning on the Public Works Loan Board. Has any analysis been done on how much would have been saved over the past year, three years or five years if more borrowing had been allowed through the Public Works Loan Board?

It is worth remembering that the Scottish Government’s borrowing powers are very new. It had very limited borrowing powers under the Scotland Act 2012 and had no such powers before that. The borrowing powers increased somewhat under the Scotland Act 2016, but they are still limited. There is a cap of about £3 billion in total, which has been a constraint.

In that context, and given how quickly the situation has changed, the Government has focused on how it can increase the amount of investment by using privately financed investment, which is what our report is about. There is clearly a live debate about whether Scotland’s financial powers should be extended further and what the borrowing limits ought to be. I am not aware of any analysis having been done on what the savings would have been. That is probably a question for the Government.

I want to look at value-for-money assessments. Paragraph 30 of the report says:

“it is ... the responsibility of individual councils and ... public sector bodies to assess the value for ... money of projects”,

and that such bodies’ assessments tend to focus

“on the benefits to be delivered”

rather than on

“the costs incurred.”

Do councils in particular have the skills to work out those issues effectively in such complex transactions?

On the broader point about the scope for councils to make evaluations, a fundamental feature of how the system works is that the Scottish Government decides where it will offer private finance support and then councils are faced with a choice about whether to accept that support or whether to undertake the work through other means. Such support is subsidised by the Government—it ranges from 50 per cent up to two thirds and, in some cases, up to 100 per cent.

We have said that, generally, councils focus on whether they can afford a project and whether it is the right thing to build, rather than on whether other funding arrangements are available. Councils have not had to call on such skills at that level.

When we spoke to councils more generally for the report, we found that larger councils had more capacity to take on some of this stuff themselves, including the procurement function, and there was frustration that the process did not enable them to do that, whereas some of the smaller councils that had less capacity were more supportive of the SFT’s role and the support that was made available in the process.

Paragraph 31 of the report says:

“affordability, rather than value for money, becomes the focus”,

which leads on from what you are saying. On the one hand, councils are focusing on the benefits rather than on the costs, but, on the other hand, they are looking at affordability rather than at value for money. That does not seem to be a great approach.

Our main concern about that is that, although we recognise the deals that councils have been presented with, there is the issue of their going into a deal with their eyes open and being clear about what the constraints are in the models that they are signing up to and the extent to which those might, for example, restrict their flexibility during the lifetime of the contract or put on-going restrictions on how some of the assets might be accessed or used. We are clear that councils need to do more in that regard and be clearer about their judgments. I entirely understand that, on many occasions, the deal is almost too good to refuse.

Who does the value-for-money assessment of the project, and at what point?

The easiest way is to think of it as having three tiers. The programme-level assessment is undertaken by the Scottish Government. That is when it decides whether private finance support should be used. In paragraphs 30 and 31 of the report, we made the point that, in the wider options appraisal in the business cases for individual projects, there is no questioning about whether capital funding or a private contract will be used—we simply see variations in the options that are presented on the basis of using a privately financed contract, for the reasons that Mark Taylor has just outlined. We will see wider value-for-money measures relating to design and community benefits, but the initial decision on whether to use capital funding or use one of those contracts is not undertaken by the individual body.

In effect, that is done by the Government.

That is right. We think that it is not clear enough on how it decides which projects attract private financing and which projects attract public capital support in the way that Mr Neil asked about.

Okay. I will move on. Paragraph 56 of the report deals with

“The age and status of NPD contracts”

and the secondary market. When I hear the term “secondary market”, I am always curious about how that market operates. Given the fact that, across the whole of the UK, a vast number of public projects of various kinds are going on, a lot of projects could be sold on the secondary market. Is there a formal secondary market? Are things done by contact between the different companies? Is there a clearing house for that?

To put it simply, it is an equity trade. It is the same as buying shares in anything else.

So shares in the company are sold.

Institutional investors will sell their shares in the company, which means that they will sell their rights to the future payments that are associated with their investment. They are bundled and sold as a whole. Ultimately, the transaction is no different from any other financial transaction between large institutional investors.

Let us say that company A owns the shares, has the contract and does all the maintenance in the buildings and so on. If it sold the shares, does that necessarily mean that it will lose the maintenance contracts, for example, or will it simply lose the right to the profit that comes on top?

This is about the structure. A special purpose vehicle will be set up, and that will remain extant. The ownership of that vehicle can move, and it might have contracted to a facilities management company. Contractors might be involved, but, ultimately, such a trade means that company A will sell its shares to company B, which will own them and have the right to a commensurate share of future payments from the public sector.

The rights to the income stream will transfer with the shares, but the special purpose vehicle will remain in place and be responsible either for directly providing the services that are involved or for managing the contracts that provide those services.

Okay. I will move on to a totally different issue. Paragraph 81 of the report comments on the Hub Community Foundation. It says that

“charities will benefit from the establishment of the HCF”,

but that that will reduce the public sector’s right to invest. That will result in reduced returns to the public sector, because the level of public sector control will be reduced with the equity balance changing. According to the report, the HCF has £3.2 million available to support charities. Do we know how much has been paid out?

We cannot answer that question. The way in which the Hub Community Foundation functions is an example of how the design of the funding schemes has been adjusted as the rules on how things are accounted for have changed. The foundation was created in response to those changes.

We found that the changes result in increased costs—they add another layer, the foundation is not able to borrow at the same rate as the public sector, it takes on more risk and therefore the debt costs more.

We do not have detailed information about the activities of the foundation beyond its contribution to the way in which the private finance model operated.

The report says that

“the public sector participant’s right to invest in sub-debt is reduced from 30 per cent of the total sub-debt to ten per cent”.

That, obviously, will have an impact on the costs. Do you have any example of the financial implications of that and how that would have impacted?

The driver for that was the change in the national accounts rules that came from Eurostat—the European statistics authority—and that the Office for National Statistics has applied. The intention of the changes was to reduce public sector control, and it has done that.

We undertook to get back to Alex Neil with information on the internal rate of return, which is mentioned in paragraph 48 of the report.

Essentially, payments are shared out. Say a payment of £1 million was made. Previously, the public sector would have got £300,000; now it would get £100,000. Therefore, the net cost of delivering that project would increase by £200,000.

If you are comfortable with this approach, we can run a model to quantify exactly what the implications of the changes would be.

That would be good. It would help us to get a grip of what additional costs have been taken on board.

Under the new mutual investment model, councils will need to fund the construction costs of new schools, while the Scottish Government will pay for maintenance and upkeep, linked to specific outcomes. An example of that is energy efficiency, according to paragraph 91 of the report. It is quite complex to measure that against outcomes. How effective will it be?

It is complex, and I am sorry to add to the complexity, but that is not MIM; it is the learning estate investment programme—LEIP—for schools. Graeme Greenhill is our expert on both those models, so I ask him to respond.

As Caroline said, the process that you outlined is the new proposed model for funding the construction and maintenance of schools. The idea is that the Government will pay revenue grants of up to 50 per cent of the capital cost of the asset. The council will need to decide how it funds the rest of the asset, which it can do by using its own sources or by taking on additional borrowing.

The Scottish Government will make a contribution towards facilities management and life-cycle costs on the condition that the council maintains the school up to a particular standard. If the Government’s contribution for facilities management and life-cycle costs is less than 50 per cent of the capital cost, it will make an additional payment that will be dependent on the council delivering certain outcomes. At the moment, those outcomes are based on things such as energy efficiency and carbon reduction—they are to do with the design of the building, in other words.

The model is at an early stage. A number of pilot projects have been proposed, but it is very new and we do not know how it will operate in practice.

The big question is always about who carries the risk. If, for example, maintenance costs are higher than anticipated, would it be the council or the Government that bears that risk?

It would be the council.

Okay. Do councils factor that in adequately? It seems to me that there is always the propensity for costs to rise.

09:30  

Because the Scottish Government’s funding is conditional on achieving a certain level of maintenance, the council would need to ensure that maintenance is kept up to a certain predetermined quality. If that requires additional costs to be achieved, the council would have to factor that into its planning.

That sounds complicated.

I will follow on from where my colleague left off. A reduction in capital funding for local government will have budget implications. How will that impact the schools programme and local government in the context of what has just been discussed?

You are asking me to postulate and hypothesise. Clearly, if there is a reduction in council capital budgets, councils will need to consider how they afford infrastructure and investment going forward. That might be by making using of their reserves; equally, it might be by borrowing more, in accordance with the prudential code.

Has any analysis done of the scale of profiteering from the selling of secondary debt under the NPD model?

The point that we are making is that neither the Government nor the Scottish Futures Trust is currently monitoring the secondary market. That would be an important first step, which would send an important signal, whether or not there are excessive profits being taken, about the appetite for that debt. It would be useful to monitor that both for current projects and for thinking about future projects. That is the main recommendation.

So is the answer to my question that you do not know the scale at all, or do you have a guesstimate?

There is some monitoring at a local level, which Martin McLauchlan can tell you more about.

There has certainly been a high level of academic interest in that research topic. It is known from PFI contracts that it tends to be after the asset has been completed and is operational—between six and 12 years after that point—that such trades start to happen.

When we undertook this analysis—we will check and come back to you on this—there had not been any secondary market trading for NPD projects. There had started to be some trading within the company groups and so on within the hub programme, both in shares in the hub companies and in shares in the individual companies that were set up for each contract. It was limited, but that would be expected. The point of our recommendation is that, as that market is emerging, it would be a good time, for all the reasons that have been touched on today, to look at how it could be monitored going forward.

Has the Government accepted that recommendation, or do we not have an answer to that yet?

We have not had a formal acceptance of the recommendation—that normally happens through the committee anyway. It is fair to say that there is a difference in view on how valuable it would be, but we set out in paragraph 58 why we think that it is important.

Excellent. There has been a lot of talk about the cost versus the value of the asset. Is £9 billion the current estimated value of the assets, or is it what we think the value will be when we get to 2048?

When the contracts are established, there is a clear articulation of what the capital cost of building the asset is at that point—that is the value that is captured in the contracts and what we have reflected in the analysis that we have done. It is the cost at the point of the asset being built.

Is there a chance that the value could be higher than that by 2048?

There is a valuation process and a Government accounting process and all the assets will be revalued on an on-going basis as part of that accounting.

At the moment, it looks as though assets worth £9 billion will cost about £40 billion.

That is right, with the caveats that we mention in the report. The difference between the £9 billion and the £40 billion is accounted for by the interest that is paid over the life of the contracts and by the maintenance costs for the assets, which are included, and, in some cases, for other services that are also provided. There is a difference, but it is not profit in the terms that Mr Neil was asking about.

It includes maintenance costs for example.

Yes.

Is it possible to analyse the hard value of the assets and what the cost of the maintenance in normal circumstances for a publicly funded project would be and compare that against the PFI, NPD or hub schemes?

That is not a straightforward analysis, not least because public sector assets are not always maintained well across their lives. When money is tight, it is often easier to reduce maintenance budgets than service delivery budgets. This committee has expressed concern in the past about the annual estates report for hospitals and national health service buildings. One of the potential benefits of using privately financed contracts is building in those costs explicitly at the beginning. That is not to say that they might not cost you more than it would by doing it to the same standard in-house, but the costs are priced in at beginning, and that is the part of the difference in the prices that we see here.

Excellent.

Is there any analysis of the £40 billion cost of the £9 billion-worth of assets by year and by different Parliaments?

I am sure that it is not possible to do any analysis by different Parliaments. I think that there is a section in the report that separates the figures out by the different types of contracts.

Exhibit 4 lays out the unitary charges for the different contract types from 1998-99 through to 2047-48.

In the publicly available information, you see the unitary charge, but you cannot dig into the components of that without going out into the models. For the nine contracts that we have looked at, we have clearly seen that the monthly cost is made up of X for interest, Y for finance and Z for services.

At a global level, exhibit 4 shows you how much is being paid for the contracts by type and by tenure.

Is it possible to compare this Government’s contracts against those of previous Governments?

You would be able to do that from the financial close date—that is, the date on which the contracts were signed. I would add a note of caution about that, because projects of this nature are large, complex and in development for a long time. However, you could do something crude and identify how many contracts were signed from X date to Y date.

I refer you to exhibit 1, which shows a timeline of all the contracts. We show the period during which PFI contracts were used, which runs up to 2009; the period in which the pre-pipeline and the pipeline NPD contracts were done, which runs up to 2017; and the privately financed hub contracts that run alongside that. It is not a straightforward handover between one Government and another. We are talking about a long period that goes back to 1992, which predates the start of this Parliament. Therefore, the analysis about which you are asking is complicated. I understand why that is of interest you, but there is not a straightforward cut-off line for us, for the reasons that Martin McLauchlan has outlined.

It is the job of the Public Audit and Post-legislative Scrutiny Committee and, more important, it is the job of Audit Scotland to make sure that we have value for money and the best delivery that we can get when using the public pound. How would you mark the schemes in value-for-money terms and what the public gets back for the scale of investment that the Government—or the taxpayer—is making?

I return to the grounds that I covered in my introductory comments. At a high level, the Government is explicit that it has used the contracts to increase the amount of infrastructure investment that it can make. There are still constraints on the amount that it can borrow or the capital grant that it has available to it. That has always been the case, and I think that that was one of the rationales for the introduction of private financing back in 1992.

The extra costs that are associated with the contracts are important and need to be transparent—I hope that our report makes a contribution to that. The other side of the equation is the benefits that we get, and we can point to the hospitals, schools and so on that would otherwise not have been built. There is no doubt that additional investment is going on.

The more important thing, I think, is the wider economic benefits that are being achieved and the reason why the particular mix of financing has been used for the particular range of projects. That takes me back to the ground that Mr Neil covered. There is no doubt that we have seen extra investment, at a cost. However, the Government is not capturing enough information to demonstrate that it represents value for money. We would like to see that step being taken as the new models are developed for the future.

Alex Neil touched on the Scottish national investment bank. Can you confirm whether it will be able to make investments in infrastructure?

I do not think that doing that directly is intended to be a major part of the bank’s mission, but Mark Taylor is leading for us on the development of the bank.

The plan is for the bank to be funded from financial transactions—from UK Government money—and there are restrictions on how that money can be used. Essentially, it needs to be used broadly to support private sector projects. The private sector does infrastructure, too, but not public infrastructure. As things stand, that will restrict the investment bank’s ability to invest in such projects.

There was a caveat in your use of the word “broadly”. In crude terms, will the investment bank be able to do public infrastructure?

We do not expect that it will be able to do that under the current set-up.

Not directly.

Thank you.

Are there funding models that are not available to the Scottish Government, or is the issue more about limitations on the amounts that can be borrowed in the markets? For example, let us suppose that Scotland wanted to build a high speed 3 railway line. We know that it is estimated that HS2 will cost £100 billion. There is no limit to the amount that the UK Government can borrow to fund such programmes, but there is a limit in Scotland. Are there limits on the models that we can use, or is it more about limits on the amounts of financing that we can deploy under the fiscal framework arrangements and rules?

It is about limits on the amounts that we can deploy. Until quite recently, the Scottish Government could not borrow at all. It had capital grants from the UK Government as part of the block grant arrangement and it could use privately financed investment in the ways we have seen. It has done that since the Scottish Parliament was established in 1999. Under the Scotland Act 2012, some limited borrowing powers were introduced, and they were increased under the Scotland Act 2016. I think that the limits are now about £475 million a year up to a total of £3.5 billion. The amount is limited in the context of the scale that we are talking about. I think that that will be worth reviewing as part of the fiscal framework review that is due in 2021, which is not a long time from now. It needs to be seen in the round as part of the package of measures that is available to the Government to stimulate economic growth and put in place taxes and tax rates that have a sustainable set of public finances around them.

Are the constraints in releasing the limits and making the powers more flexible basically political constraints?

There are political constraints, but there are also constraints around the extent to which that package as a whole works. The UK Government has control over all the levers for both raising and spending money. The Scottish Government does not have that. The UK Government has control over important economic levers such as immigration, which the Scottish Government does not have. We need to think about the overall package of powers and then about the way in which the more limited fiscal powers fit within that and how we can ensure that they are consistent and sustainable over time. The detail of that is a political question that is way above my pay grade, but there is a very important interaction between the borrowing powers, the wider fiscal framework and the broader devolved settlement.

What is the biggest capital project that has been undertaken in Scotland? Is it the Queensferry crossing?

The Queensferry crossing was funded from the capital budget.

Was the cost about £1 billion?

It was a bit more than that.

Does that show the limitation on the powers that the Scottish Government has to deploy? Is it about that figure for a single project?

That is a choice for the Government. It controls its infrastructure programme and pipeline and it could prioritise a smaller number of large projects rather than a larger number of small ones. However, it is constrained by the package of the capital grant, the borrowing powers and its private financing capacity.

If a major project emerged that could cost £4 billion, would it be a non-starter, given the borrowing ability?

I do not think that we have looked at the overall picture of how it could add all those things up, but there would be a limit to how much it could afford to invest.

09:45  

To add a wee bit of colour to that, there is a limit on the overall capacity. Another example is the dualling of the A9, which the Government is planning. I think that it is expected that the overall cost of that will be around £3 billion, which will be spread over a number of years. There are limits on the Scottish Government’s capacity to fund capital investment in any one year, and the design of the finance private schemes is to augment that. However, things can be built over a number of years. How best to do that is a political decision.

Thanks.

I want to ask a couple of questions about the role of councils, which colleagues led on earlier. Paragraph 32 of the report says:

“In deciding whether to pursue and accept ... Scottish Government funding, it is important that councils and other public sector bodies consider the extent of future financial commitments carefully”.

You have talked about the role of councils and whether they are fully participating in decision making. Will you tell us a wee bit more about that and whether that role has been fully deployed? Can councils’ role and participation in taking projects forward be improved?

As we said in answering an earlier question, the starting point is that it is very hard for the council to discount the Government’s offer to meet part of the costs. That is a free good. If I offered to pay half of your mortgage, you would be more likely to look at what you could afford with the other half than at the value for money of the overall house that you are buying.

Earlier, we mentioned the prudential code, which binds council borrowing. Basically, it says that councils should not borrow more than they can afford to repay. How much each council can afford to repay is very much for it to determine, and that will differ between councils because of their reserves, commitments and existing borrowing. Earlier, we spoke about the 5 per cent cap that the Scottish Government has set for itself in respect of how much arising from borrowing and unitary charges associated with private finance it is willing to commit to repaying. I am not aware that individual councils have set a similar percentage figure, but it is clear that they have to think about the affordability of future repayments in setting their own borrowing limits.

Of course. Is the prudential code more flexible in respect of the cost of borrowing? Is that more effective than the Scottish Government’s powers through the models that we are talking about?

It is more flexible, but it is difficult to say that it is more effective.

I should have said more cost effective.

The prudential code gives councils more flexibility and, used well, it will allow the sort of decisions that you are talking about so that they can say that it is worth their investing the additional amount because they think that the wider benefits will justify that. However, it needs to be used with care. We know that budgets are under pressure and that there can be pressure to borrow more than it might be prudent to borrow in the longer term. It also introduces another element of risk. Most councils borrow from the Public Works Loan Board. There was quite a significant increase in the borrowing rates just before Christmas, which will have a knock-on effect on borrowing for years to come. That sort of risk has to be taken into account, as well.

We know that councils could not borrow billions of pounds, but their mechanisms to borrow are more effective than the Government’s powers.

They are more flexible.

That is incredible.

I have a final question about the role of councils. Councils have a role to play in the long-term repayment of the loans, but can they accurately estimate the costs of the repayments over the period? How does that impact on their delivery of local services? They set aside repayments for programmes. I presume that, if some other means of funding were available, that money could be set aside to deliver local services.

The whole principle behind private finance is that it gives councils certainty about the unitary charges that they will be required to pay over 25 years, or whatever length of time the contract is for. That has to be built into their forward planning. The process involves committing money for the future, which inevitably reduces flexibility in how councils use the rest of their budgets.

To give a local example, in Ayrshire, we have a residual PFI in Ayrshire College, on which there are still six or seven years to pay. However, in another example, the Scottish Funding Council bought out a PFI contract. Is it possible or cost effective to buy out PFI contracts? That has happened with West Lothian College, but there is an on-going PFI in Ayrshire College.

It is generally possible, depending on the contract that was put in place. Whether it is value for money depends on the negotiation between the public body and the special purpose vehicle. On the whole, if the present value of the returns to the special purpose vehicle and its investors is more than the price that would be paid, it will not be possible to buy out the contract without paying more than would be value for money. That is a complicated way of putting it. There would be a negotiation with a set of private investors, who would look at the payment that is currently on offer versus the stream of payments to which they are entitled under the contract, which means that buying out the contract is likely to be more expensive than continuing to make the payments.

I have questions on a couple of areas. In paragraph 52, you comment that the Scottish Futures Trust

“does not disaggregate its reported benefits”,

which

“makes it difficult for councils and other public sector bodies to establish and report the savings delivered by making use of centralised approaches.”

You recommend that the Scottish Futures Trust should

“demonstrate more clearly the links between the financial savings attributed to centralised activity and individual projects”.

Is there any explanation as to why the Futures Trust does not provide that disaggregated information?

No. That is the approach that was agreed with the Government when the Futures Trust was set up. Martin McLauchlan may want to add something.

The benefits are calculated by looking at overall expenditure and applying an effectiveness factor, which takes into account not only the individual project work that has been undertaken but, for example, the savings that could be attributed to using standardised documentation and to other innovations that might be made centrally. If you apply 10 per cent, which reflects X, Y and Z, it is difficult to break that down. When we asked people who are involved in the projects whether there has been a 10 per cent saving, we found that they have no comparator in that regard. That has, in some ways, made it more difficult to break down the figures into a meaningful contribution.

Are you saying that the Futures Trust could disaggregate to a certain level and would then have to allocate the rest on some basis?

The figure is calculated on the total spend, which is actual expenditure. If there is spend of £100 million and you know that £10 million of that has been spent on an individual project, surely it follows that 10 per cent of the reported benefit relates to that individual project.

By “surely”, do you mean that it follows or that it does not follow? Are you making that assumption?

The Futures Trust makes that assumption. Our point is that it should be clearer about what the benefits are. The key point is that the Futures Trust is trying to estimate against a counterfactual that would not exist, because it is involved in the only private finance deals that are available. The Futures Trust is trying to say that, because it is involved and there is standardisation, the private deals that it is involved with are a certain amount cheaper than other private deals would be. We are somewhat sceptical about the calculation methodology.

I admit that I do not know a huge amount about the Scottish Futures Trust. Who regulates it?

It is a company that is wholly owned by the Scottish Government. On page 12, the report sets out the situation. At the moment, the company appoints its own auditor, but there are proposals to move it to the remit under which the Auditor General will appoint its auditor in the future.

Do you mean the financial auditor?

Yes.

My question was really about who regulates the Futures Trust.

It is not regulated. It does not provide functions that would normally be regulated. It does not provide finance; it operates as a facilitator for private financing and private contractors.

If it is not regulated, who monitors it and holds it to account?

The Scottish Government—it is wholly owned by the Scottish Government.

In practical terms, what does that mean?

I think that part of the Government’s director general Scottish exchequer team deals with the SFT. There is an infrastructure team that has a relationship with the SFT.

Do you know what goes on in the Scottish Futures Trust?

We looked at the SFT as part of the audit. As always, we have rights of access in the normal way. We are clear that it has delivered some real benefits by providing expertise and a critical mass in relation to developing the standard documentation and so on, which Martin McLauchlan has talked about, and carrying out the market exercises for financing. We are saying that it is not clear enough how those benefits relate to individual projects, rather than there being a benchmark saving that is supplied across the whole programme.

At paragraph 96, you make a number of recommendations about transparency and note that better information is required. What specific information about the new investment models is required in order to correct the weaknesses that you have identified in relation to the transparency of the NPD and hub projects?

I will pull out a couple of key things. As we discussed earlier, there would be real value in monitoring the performance of contracts in financial terms after the initial signing—for example, in relation to the secondary market, if that is still relevant.

More generally, as we say in paragraph 97,

“there is a need for a more regular assessment of the contribution that ... private finance is making”

to the Government’s wider goals and, in particular, to the wider economic benefits that come through. The infrastructure mission and the infrastructure investment plan that is due later this year are based on the assumption, for which there is good economic evidence, that there are wider economic benefits to investing in infrastructure. However, more could be done to genuinely evaluate and monitor the benefits to ensure that the right mix of projects and the right mix of financing are being used.

You mentioned the secondary market. There is nothing illegal in the secondary market, is there?

Not at all.

It does not change any of the payments for the project to the special purpose vehicle.

No, but, as we say in the report, it sends important signals about the contracts. We think that there would be value in the Scottish Government and the Scottish Futures Trust having that information.

I understand that.

The report’s first recommendation is that the Scottish Government should

“better document and report how decisions on the use of private finance are made at a programme level, and how the overall combination of programme and project funding aims to maximise investment and benefits”.

You say “better document and report”. When you were looking at and asking about the issue, did you find that the Scottish Government just could not write the information down, or could it not tell you how it had made decisions?

Understanding of what is most suitable for private finance has got better through time—we have set out the timeline from 1992 to the present date. Through our discussion with the Government and the SFT, we know that they have a better understanding of which projects are most suitable for private finance and some of the characteristics of such projects. We capture some of that information in our report. Although they could articulate, to an extent, why certain projects had been subject to private finance, there was no clearly articulated trail of documentation to show how individual decisions had been made. As important—or more important—it has not been clear to the public and the Parliament why certain projects have been chosen for private finance and others have not.

I am thinking about the practicalities. The documentation, if it exists, could be from 15 years ago.

Some of it could be.

Where would such information be kept so that it can be found when you look for it?

The Government keeps records over a long period, as you would expect. A lot of those records end up in the National Records of Scotland, and it has not been unknown for us to go back over long periods to look for information. As Mark Taylor said, we are looking for the Government to demonstrate that it has a clearer picture and for it to be accountable to Parliament for how it makes the choices that Willie Coffey and Alex Neil touched on. How is the Government using the capital grant, its borrowing powers and the private financing that is available? The Government should look at the mix of that and whether it is worth doing more, and it should monitor whether it is getting the benefits that it expects to get.

Out of interest, when you were asking those questions, did anyone have to delve into the archives to find stuff?

Given the conclusions that we are drawing, it is fair to say that a lot of efforts were made to give us additional evidence, but no such information on specific decisions was forthcoming.

There was nothing to dust off.

10:00  

Auditor General, in your response to Willie Coffey, you talked about buying out PFI projects. When the Scottish National Party Government came into power, it abolished car-parking charges in national health service hospitals across the country, except for three, which were run under PFI contracts. Do I take it from your answer to Mr Coffey that, if the Government were to buy those hospitals out, the question of whether that would represent value for money would be dependent on negotiation with the investors?

In broad terms, yes. Every privately financed contract, whether it is PFI or NPD, will include some services that are bundled up with construction and maintenance. Those services tended to be more extensive under PFI projects and included things such as car parking. If the Government or the public body that is involved wants to change those services, it needs to negotiate that with the special purpose vehicle—the company that has been set up to deliver the services. In most cases, it will be possible to do that. However, it can be done only at a price that is agreed with the company itself, and the Government or public body might conclude that it is simply too expensive to do that, given the benefits that would be achieved.

There might be circumstances in which it is not possible to do that, but I think that whether it is done or not is more likely to be decided by the result of a negotiation and a consideration of whether it demonstrates value for money.

Do you have anything further to say on the issue, Bill? I know that you have been working on it.

When I read the contracts, I saw that they are very long and complicated. However, I think that they contain mechanisms that would allow them to be bought out. As is the case with any commercial negotiation, you must have two sides that are willing to tango.

I forgot to ask something on the issue of the secondary market models. You said that the secondary market functions on the basis that the shares in the SPV are being sold. That is fine for PFI, but it does not work for NPD, as I understand the situation. Paragraph 54 of your report states that

“companies ... generate profits by trading their rights to future contract payments to others.”

What sort of instrument are they using to do that? Clearly, from what you are saying, it is not shares. However, they must be wrapping those rights in something so that they are able to sell them on.

As Martin McLauchlan said, because of the age of NPD projects, we have not yet seen that process happening with them. I do not know whether we know anything more about what vehicles are being discussed in relation to the issue. Martin McLauchlan is better placed than I am to answer that question.

I understand that the future payments are apportioned to shareholders: owning shares gives them the right to a share of the future contractual payments. The price at which those shares are sold will reflect the present value of discounting. If you know that you are going to be paid £1 million a year for 25 years, you are not going to—

You are saying that it is actually the same model, although you have sort of packaged it in a different way in the report—is that correct? It still comes down to the SPV shares being sold.

The membership of the company is different, but, ultimately, the situation with regard to the SPV is no different from what happens in relation to a listed company selling shares.

The difference concerns the right to dividends from companies under PFI. We set that out in paragraph 54. PFI entitled the shareholders to a stream of dividends, and, as we all know, the value of dividends can go up and down over time. In relation to some of the earlier PFI contracts, there were records of genuinely excessive profits leading to dividends and significant secondary transactions. That right has been removed in the NPD model, but the special purpose vehicle is still made up of a group of companies that own shares in it, and they can sell their shares in that company to others. Those shares might not be sold on the open equity markets—they might be private transactions—but the share in the company brings with it the right to the payments that are made, as well as the costs that are associated with the contracts.

I have a couple of questions. On that point, is there any evidence of offshore companies being involved, either as senior debt players or junior debt players or in terms of the reselling of the equity in the special purpose vehicle?

In paragraph 50, you refer to what I think it is fair to describe as excessive professional fees that, in some cases, amount to 4.5 per cent of the capital value of the project. Is that not another rip-off? Can you give us a breakdown of the professional fees in the three cases that you refer to in paragraph 50?

In paragraph 50, we talk about “significant” rather than “excessive” fees.

Yes, but I am saying that fees of 4.5 per cent are excessive. It is a nonsense.

I think that you attributed the word “excessive” to us, but we use the word “significant”.

You are too modest.

Thank you. We will look at what breakdown of those fees we can provide.

On the issue of offshore ownership of some of the shares, we know that some of the equity is held offshore. Again, I think that almost all of that will relate to PFI contracts, because they are older and there has been much more of that sort of secondary work in relation to them.

Would you be able to give us a list of offshore companies that you know of that hold those shares?

I do not think so. The data that the Treasury and the Scottish Government publish shows the make-up of the SPV, but the ultimate corporate ownership of that investor would be outwith our remit and expertise.

We will have a look and see what we can provide.

If you cannot provide the information, could the Scottish Government do so?

We think that it is not monitoring that at a programme level—that is one of our findings. However, as Martin McLauchlan said, there is academic research and campaigning research that we can point you towards, if that would be useful.

Thank you.

Should there be monitoring at that level?

One of our recommendations is that monitoring the secondary market would provide valuable signals that would help to inform future contracts.

As we have no further questions, I thank you for your evidence this morning and close the public session of this meeting.

10:06 Meeting continued in private until 10:30.