“Management of the Scottish Government’s capital investment programme”
Item 3 is consideration of a section 23 report, “Management of the Scottish Government’s capital investment programme”. I invite the Auditor General for Scotland to brief the committee.
Thank you, convener.
Thank you very much, Mr Black.
The previous Administration used the private finance initiative significantly. Exhibit 9 on page 16 of the report highlights how important that was. It describes the capital value of PFI and non-profit-distributing projects since devolution, by sector. Alternatives to PFI have been under development for a number of years. That started under the previous Administration. The team might be able to help with more information about that earlier stage.
I think that you have captured it well; I am not sure that there is much more to say. The NPD initiative was something that came in in the early days of the Scottish National Party Administration.
Is it not the case that a couple of NPD models were developed prior to 2007?
There was a pilot project in the schools sector, which was progressed under the Labour Administration as an experiment. Greater priority was given to NPD under the SNP Administration. Indeed, a policy decision was reached that, in future, it would be the preferred form of finance.
The previous Administration described it as PPP. You distinguish between PPP and PFI. What is the difference?
I think that it is fair to say that PPP—public-private partnership—is the umbrella phrase that applies to all such projects. Various non-profit-distributing vehicles and PFI vehicles come under the PPP umbrella and are used as appropriate.
So the non-profit-distributing model is essentially still a form of PPP.
Yes, the essence of that model is that there is a profit cap or that, after the financier and the developer have made a return, the profits come back into the public purse.
Okay.
Yes.
Yes.
I will follow up that line of questioning. Exhibit 7 on page 13 of the report gives the definitions of the different methods of financing. In reference to the non-profit distributing method, it states:
That is a reasonable way to refer to it in many ways.
We have not audited any individual NPD projects, although we have looked at a number of PFI projects in the past. My understanding of NPD contracts is that the company that is formed to provide the services that are required under the contract is unable to distribute profits. The profits are retained and may be applied to public projects, but they are not distributable.
Does that apply to all the profits? There is a suggestion, in the definition, that a certain amount of profit is retained by the funder and it is the surplus profits that are—
Yes. As I commented earlier, the funder clearly requires to make a return on their investment. However, a cap is placed on that and any surplus is returned or shared with the other parties.
We have tried to use high-level language. There are obviously profits in any project, as any project involves engagement with the private sector. Therefore, a construction company that carries out work under an NPD contract will earn profits from that activity. The point of an NPD arrangement is that it restricts what the special-purpose company that is set up as the focus of the project can do with its profits. It cannot distribute profits to its shareholders—that is the essence of the arrangement.
Just let me clarify that again. Murdo Fraser is saying that there will be profits—you have confirmed that—but they will be capped.
Yes.
So, there are profits, but any surplus profits—whether there are any remains to be seen—do not get distributed.
If it would be helpful, we could provide a note to the committee on that. I am having to use terms such as “special-purpose company” because it is quite a technical area legally and contractually. You have got the gist of it. If you want a bit more specific information, we can provide it.
Yes. We just need to get our heads around the fact that the companies that are involved in the projects will still make profits, so it is a form of PPP. Also, just as previously, the public sector will pay an annual charge over the life of the asset—there is no difference there.
It is exactly the same arrangement.
In the period ahead, budgets will be challenged, so it is important that good monitoring arrangements are in place to track money that is due to be spent on capital projects and money that is forecast to be spent in the future. It is a matter of concern that, in paragraph 16 of your key messages report, you state that, in relation to the revenue charges that result from capital investment, you have had difficulty in getting information for some of the projects—especially the traditionally funded projects—regarding on-going revenue costs, which makes forecasting difficult. In the case of depreciation, for example, standard policies are normally applied and, if the initial capital investment is known, it is relatively straightforward to project the depreciation going forward. However, your report underlines the fact that, for certain projects, you have been unable to get on-going information relating to depreciation charges.
That is correct. We could not provide in the report high-level information on those on-going cost requirements of depreciation, running costs and so on. I am sure that the team can provide a bit more detail about the problems that they faced in addressing that issue.
The issue is about strategic—forgive me for using that word—information that is available to the Government. It is perfectly possible, given enough time and resource, to identify the forecast running costs of an individual project.
I appreciate the challenges that you face as a small audit team, but I am trying to establish what the Government does when it is trying to forecast investment that is to be made. Correct me if I am wrong, but I interpreted paragraph 16 in the key messages report to mean that the information was not available—either from the Government or from—
That is absolutely correct, and it is one of the reasons why we have suggested that the Government might wish to consider revising or revisiting its infrastructure investment plan, so as to make that a more strategic document. The point that you have highlighted is one of the issues that the Government should be considering.
Okay. I reiterate the importance of having accurate information, and the Government has to take that on board seriously.
I can understand the reasoning behind that concern, but it has been very common for years, in managing a large capital programme with many projects, to follow such an approach. I recall from my previous life as the chief executive of Tayside Regional Council, when I chaired the financial planning group, considering the projects that we intended to have in train in any one year. In the course of that year we would almost inevitably encounter slippage somewhere. That is just life when major capital projects are concerned.
We try to write our reports in plain language, and we have perhaps failed in the paragraph to which Mr Kelly referred. The £100 million of overallocation each year is for revenue and capital—it refers to the whole of the Scottish budget, so there has not been a £100 million overallocation just for capital expenditure. It is intended to allow some flexibility between revenue and capital and, as the Auditor General said, there will inevitably be some slippage, which allows the £32 billion to come in closer to the mark.
How is that reflected in the budget documentation? I do not recall seeing it in any previous budget documents.
The purpose of the approach is to help the Government achieve the budget that has been set. These are the Government’s spending plans against the budget. You may recall that we picked up on this issue in our report “Scotland’s public finances: preparing for the future”. More recently, the committee considered the outturn of the Scottish Government’s consolidated accounts where subsequent evidence came in on the actual outturns over the 10 years since devolution that showed that, because of the practice we are discussing, the outturn on the overall budget was smaller than it had been in previous years.
I understand your point, but it was certainly news to me that, in effect, £100 million of slack had been allocated. It might have been in the documentation, but I did not pick it up in previous budget documents or in the current budget.
I would not encourage the committee to use the term “slack”, because the performance is quite clear: at the year-end, the Government comes in very close to budget. The £100 million is essentially a financial management tool that is used in the Scottish Government to try to ensure that that happens.
Just to be clear on this, are we saying that the Scottish Government budgets to spend £100 million more than the budget and that that is not disclosed in any documents?
The recent spending reviews have specifically stated that there would be a £100 million overallocation each year compared with the budget that was being approved. That is just the spending plans.
Yes, I understand. We are talking about spending plans that are budget plans. Is that £100 million disclosed anywhere?
It is in the spending plans and, as far as I am aware, it will be in the budget.
It is included as a description.
In the budget document, do we see a budget figure plus £100 million? How is it described? How do we get transparency, openness and disclosure on this issue?
You read the budget.
Where is the £100 million allocated?
It is in the budget, but it is allocated over hundreds or thousands of individual projects.
The allocation takes place at a level that I do not think is disclosed in the budget document. Allocation happens at project level and that project investment is not disclosed, because there are hundreds of projects at the same level of detail. It is not disclosed in the budget estimates. So, the budget is the budget and there is no uncertainty about that, but there is a description in the estimates of the financial management process in which the Scottish Government engages to help ensure that it spends as close as possible to its annual budget and does not overspend.
But let us say that the budget is £30 billion, in round terms. Do the budget documents show that the Scottish Government has budgeted for £30 billion plus £100 million?
No. Because the budget is the budget it shows that the figure is £30 billion, but a kind of disclosure in the budget documents states, in effect, “We have adopted this practice.”
If we were to add up the individual budget headings, they would come to £30 billion and £100 million.
Yes, if they were available. However, they are not available because—
Because they are not disclosed at that level.
Exactly.
That means that if the Scottish Government chose to budget to £30 billion plus £130 million, let us say, that would simply be disclosed and regarded as normal practice. Is that right? In other words, the figure of £100 million that the Scottish Government overbudgets is clearly not a magical figure.
It is a matter of judgment on the Government’s part.
Judgment is a good term. It is a judgment whether the figure should be £80 million, £100 million or £200 million, let us say, because of the level of underspend that is generated each year by the Scottish Government, because there has never been an overspend. Is that correct?
I do not have the figures to hand, but my recollection is that four or five years ago there was a pattern of the Scottish Government underspending against its budget by of the order of £500 million or £600 million per annum. That was seen to be a poor use of resources, because the Scottish Government was spending less money than it had.
I think that the committee has received those figures.
Yes, it has.
I do not think that the normal underspend of £500 million to £600 million that you have quoted is accurate. There has always been an underspend and there continues to be an underspend. There was an underspend last year, as I recall. My point is that there has never been an overspend by the Scottish Government. The level of the underspend has varied substantially from year to year over the devolution period. You are saying that the current policy is deliberately to overbudget by £100 million in order to try to ensure that the budget comes out closer to the actual figure, but that is not detailed in the budget documents at a level that would be clear to MSPs and this committee.
As I think that Angela Cullen said a moment ago, the language of this is really difficult.
It is very important.
The budget is the budget, approved by Parliament. However, in managing that budget within Government—with literally hundreds of different projects—the Government will allocate more than the actual budget at the margin, because there is always slippage and projects can always be reined back three quarters of the way through the year once the projected outturn has been done round about January or February for year end. To my mind, that is good financial management.
But, Mr Black, would it not be reasonable—
It is good financial management because it means that you are getting closer to the budget at the year end.
But would it not be reasonable to disclose how that is done and to be transparent about it? I am seeing absolutely no transparency on that issue.
As I think that Dick Gill mentioned, there is disclosure of the general approach in the budget.
I am sorry, but I am talking about understanding how that £100 million is allocated across the budget.
Convener, would it be helpful—
Hold on. I will let James Kelly in and then Anne McLaughlin.
I just want to make the central point that we are going through the budget process at the moment. It is certainly news to me that £100 million more than the actual budget is being allocated.
Let me be clear about this. From what I am hearing, in respect of the overall figures that are provided to members, there is a specific figure and the Government seeks to live within that figure. However, at the level of information that is not provided to members, there are budget allocations that, when totalled, come to more than the budget provision.
Yes, by a small margin.
By £100 million. So, if you were to add up the figures that we do not see, the budget is £100 million more than what is indicated to members. You are saying that that is good financial management because, in the course of the year, things will even out and the Government will stay within its budget. Is that correct?
The significance of paragraph 36 in the report is not simply to inform the committee of the financial management practice that takes place in year but to make the point that the risk of overspend might be increasing in an era of declining resources. In other words, it needs to be managed very carefully to avoid the risk of breaching the budget limit.
Until now, that might be an issue, because when members are considering the budget they do not know that level of detail—that there is an overprovision or an overanticipation. If members are looking at budget headings, they are not able to challenge where the money is being spent because they are simply being told, “Here is the top line”, and they do not see the detail. Would I be right in saying that that will become less of an issue in future years because, as we have discussed before in the committee, we are moving to approval of top-line figures rather than detailed scrutiny of departmental figures?
Yes. In essence, the budget act will be based on a single figure, within which the Scottish Government must live, to use your phrase. The Government has given an undertaking that it will not reduce—and will continue to try to improve—the quality of in-year financial reporting that it gives to the Scottish Parliament.
However, as James Kelly and Nicol Stephen said, we cannot see that level of detail even now, never mind in the future, because we do not get the breakdown of what is budgeted for individual projects. We will see only the top-line figure.
There seem to be two interpretations of the situation. Mr Black, do you think that it is something shady and underhand, which you need to investigate, or is it a perfectly legitimate and not unique accounting tool to ensure that we reduce the amount of money by which we underspend and therefore reduce the amount of money that goes away from public services in Scotland and back to the United Kingdom Treasury?
Before you answer, Mr Black, I want to put it on record that the committee is not suggesting that anything shady or underhand is going on. I would not want anyone to go away with that impression.
Not the committee, but Nicol Stephen.
There is a separate issue about transparency and openness. If Mr Black could answer in that context, that would be helpful.
The practice that has been adopted in recent years by the Scottish Government, of overallocating at the margin within the set budget, has contributed to better financial performance, as measured by the outturn against budget in recent years. That was documented in past reports that have been given to the committee.
That is perfectly clear, and it sounds as though it is a worthwhile financial management tool that is worthy of adoption. However, the issue that was being raised was not whether it was a good financial management tool but whether it was effective in terms of accountability and parliamentary scrutiny of the budget process. That is a debate that is best left to the politicians rather than you, Mr Black.
What the convener has just said is absolutely correct. It is an appropriate debate to have. The matter should perhaps not be pursued at this point, but it is something that, as politicians and as a Parliament, we might consider pursuing. Someone suggested that £100 million is a small amount of money. However, in the context of some of the debates that we have had since devolution, £100 million is a substantial amount of money. In the context of the debate that James Kelly is talking about, for example, £30 million is a significant sum of money. Further, in the debates around tuition fees in which the Liberal Democrats and Labour were engaged in 1999, we were talking about sums of money that were substantially less than £100 million.
Does James Kelly want to raise anything else?
No, I have covered my points.
Anne McLaughlin?
I might be misunderstanding what Nicol Stephen is saying. Is it the case that there is an extra £100 million floating about that could be spent on the things that Nicol Stephen is talking about? I do not think that £100 million is an insubstantial sum. Is there £100 million floating about that we are not using?
No.
To further clarify that point, is that £100 million included in the headline figures in this area of the budget?
No.
The budget figure is the budget figure. Once the budget figure has been set, the Government manages the programme by overallocating in certain areas in anticipation that there will be slippage and that—
I understand that. It was not clear to me whether it was in the headline figure.
There is no sum of £100 million concealed anywhere.
Murdo Fraser invited Mr Gill to accept the definition of NPD as a capped-profit-distribution model. In that regard, would it be fair to characterise PFI as an uncapped or unrestricted-profit-distributing model?
I cannot think of any way in which it might be restricted, other than by the contract that is entered into with the Government, as it would depend—
An indirect answer to Mr Hepburn’s question would be that PFI contracts are usually struck after a competitive process.
Yes, I mean—
After that competitive process, will the revenues be restricted to what has been contractually signed?
Yes, in essence there will be a fixed price for the contract. That will be agreed in advance, so in that sense there is a cap, in that you would expect to continue to pay the contract sum over the lifetime of the contract. What is not capped is any element of profit that the private firm may derive from its activity over the 30-year contract.
Can you think of any PFI contracts—as opposed to NPD projects, when it is part of the model—in which some of the money is reinvested back into the public sector?
We have not looked at any PFI projects in that way as part of this project. A couple of NPD projects have been completed in the schools sector, but even those would have been running for no more than three or four years and they have a 25 or 30-year contract life, so the idea of reinvesting profits at the end of the contract has not yet arisen.
So we do not know that, in any NPD contract, money will be reinvested back into the public sector. It is too early to say; it might happen, but it might not.
We do not have any evidence because, as I am sure you appreciate, the report is looking at the management of the capital programme rather than the use of finance.
Exhibit 9, on page 16 of the main document, shows the total capital value of PFI and NPD projects. The schools spend is the part of the bar that is coloured blue. There is a very substantial level of spend in 2005-06, 2006-07 and 2007-08; the figure is around £500 million in 2005-06, and it rises to almost £1 billion in 2006-07 and 2007-08. However, it collapses to under £100 million in 2008-09 and seems to disappear completely in 2009-10.
I will invite Dick Gill to come in shortly and explain the movement to the best of our ability.
I do not think that it is exclusively for the education programme.
But a large part of it is for the education programme.
Yes.
Did you say exhibit 4?
Yes. It is the second item in the table on page 10.
Is that a PFI or NPD programme?
It is a mixture of traditional finance and NPD. It is also a mixture of finance from the Scottish Government and the local government sector. Note 1 on exhibit 4 says:
Is that for the future?
It is, yes.
My question is about exhibit 9.
I am trying to say that the investment has gone in cycles and that what you see in exhibit 9 is a significant cycle of investment in the schools programme. Exhibit 4 suggests that another significant investment cycle will be coming up during the next six or seven years.
Will the £800 million be spent over the six or seven years, or is it for each year?
I am afraid that I do not have a profile of the spend with me.
Is £800 million the total?
It is, yes.
I do not wish to mislead the committee but, if I recall correctly, I think that I indicated that the sum on that line in exhibit 3 is for schools only. I apologise for that. Clearly, it will include other things such as major waste management schemes and other projects.
You were talking about a period of six or seven years.
Yes.
Over that period, and being generous, an average of £150 million a year will be spent. It looks as though well in excess of that sum was spent in the years 2004 to 2007.
The schools for the future programme has a £1.25 billion investment programme. It is spread over six or seven years, so that would be approximately £200 million per year.
I am talking about the Scottish Government’s contribution. Is what we see in exhibit 9 the total spend on each of the heads, or is it the Scottish Government’s contribution?
We are comparing two slightly different things. Forgive me; that is possibly my fault for introducing the idea. Exhibit 9 deals with NPD investment, so the £1.25 billion is likely to be a mixture of NPD and traditional finance. The situation is further complicated because some of the finance will come from the Scottish Government’s capital budget, and some will come from local government’s capital budget.
There seems to have been more than £2 billion spent on schools in the three years 2005-06, 2006-07 and 2007-08. Is that figure correct?
It is something of that order, yes.
If my reading of the exhibit is correct, it looks as though the figure is substantially more than £2 billion. In three years, £2 billion was spent. In the following two years, less than £100 million was spent. You are saying that in the next six to seven years, perhaps £1.25 billion will be spent on schools.
Yes.
Thank you.
I want to ask about some of the findings in part 2 of the report, which covers the performance of recently completed major capital projects. In previous meetings, over a long time, the committee has looked at the fairly dismal record that we have on cost and time overruns in capital projects. Indeed, if we are to believe what we have heard this morning from the BBC, your report on the Edinburgh trams, which is due to be published, suggests a similarly dismal picture—although I dare say that you will not want to comment on that until the report is published.
I will kick off, and Dick Gill and Kirsty Whyte can chip in. We highlight the fact that there were significant gaps in the information that was available to us during our review, so it involves a smaller sample than the 35 projects that were completed up to 2010. We found no difference in performance in meeting cost and time estimates between traditionally financed projects and revenue-financed projects—or they were very similar in performance.
A lot of the time delays that are noted in the report tended to occur at a very early stage. Do a lot of time delays occur before the decision is made about how a project is to be funded?
That is a fair comment. In exhibit 14, on page 24 of the report—it is the upper exhibit—we have marked with an asterisk those projects that were revenue financed. In the section for projects in the health sector, which is towards the right-hand side of the exhibit, there are significant differences between the forecast and actual completion periods. A lot of those differences are associated with revenue-financed projects. A feature of revenue finance is that it is a complex process that involves constructing contracts with a 30-year life, which introduces all sorts of complexity and challenge into the commercial arrangements. That means that it can take a long time to get it all sorted out. Some of those projects that became revenue-financed projects may have been originally designed to be progressed as capital-financed projects but the capital budget was not sufficient to allow that. Different financing methods had to be investigated, which took time, and there was slippage in the overall completion periods for the projects.
That is helpful. That deals with time delays. Can we move on to the cost estimate comparisons in exhibit 13, which, I confess, I had difficulty in deciphering?
It is quite complicated.
It looks like a game of battleships.
Can you talk us through what that graphic tells us?
Whatever you want it to.
I will start. We compare the cost estimates at two different points, which are the two benchmarks that we think should be used. The first is the point at which the project is first approved and reaches the outline business case stage; the second is around the contract stage. The first estimate is important because, when a project is started, a realistic assessment of what it is going to cost the public sector is required. That is important for appraisal and strategic decisions about choices between projects. The estimate at the contract stage is important because, if that estimate is not achieved, there is every possibility of inefficiency and waste going on—it is an indicator of poor management of the project. We looked at both those measures.
And that applies regardless of how the project was funded.
Yes. The projects in exhibit 13 are exclusively revenue-financed projects. The equivalent data for traditionally financed projects are in exhibit 11. If you wish, I can provide some commentary on that but, essentially, the cost estimating is improving compared with the position that we reported in our 2008 report.
If it helps, I will give a quick summary of exhibit 11. We appreciate that it is quite hard to digest. Sixty per cent of projects were within the initial cost estimates and 40 per cent were not, but 86 per cent came in within the contract price that was signed and only four projects did not. The combined total cost that was paid out was less than the contract price that was signed for all those projects.
Both the statistics that Angela Cullen mentioned are significantly better than the equivalent statistics that we reported for the earlier set of projects in our 2008 report. In our current report, 60 per cent of projects completed within the initial cost estimates, whereas it was just 40 per cent in the previous report. Eighty-six per cent of projects completed within the project approval price in the current report, but it was just 58 per cent last time round. There is quite strong empirical evidence that estimating is improving.
Okay. I think I understand most of that. Thank you.
Your report recommends that
I think that everybody recognises that these things should be done, but somehow they do not always get done. Again, the position is improving compared with the position in the 2008 report—we have statistics on that—but one of the difficulties that the Government has had is getting good information because of the sheer number of projects.
I have a couple of questions to round off the discussion. Do you know why the Government is moving from traditional funding to public-private partnerships for the city of Glasgow college and the Royal hospital for sick children in Edinburgh?
The short answer is no. We did not look at individual schemes in detail and we did not look at the financing. The report is simply a high-level report. I am sorry.
Okay.
We have not done that as part of an audit, but the consideration of delivery options would be essential as part of the appraisal process for any project. For a revenue-financed project, one would certainly want to compare the estimated costs using that model and the whole-life costs using another form of finance. That is a standard part of an investment appraisal.
We say in the report that the information on whole-life costs for traditionally financed projects is not completely available, so comparisons would be quite difficult.
Yes, it is difficult to make comparisons. When people criticise a model, it is difficult to make a valid comparison with another model.
That can be done at the project level, but our difficulty is at the programme level. Such information cannot be extracted without going to individual projects. With hundreds of projects, we were not able to do that.
Finally, you have outlined that there are a number of very good management proposals relating to the control of costs and reporting, for example, but I am worried about the future of capital investment that you have highlighted, irrespective of who will form the next Administration after this year’s election. It is quite clear that there are major, hugely expensive projects in the pipeline. You have indicated to us graphically the significant reduction in the availability of capital and you have highlighted at various points the reduction in revenue. Previously, you have reported to us on the pressures on the maintenance of roads and the school estate, universities and a range of other things that are not referred to in the report. I am worried about the point at which either you will give a reality check to politicians or there will be a debate on what we can realistically afford.
Yes. I am afraid that it is quite often my humble duty to tell members when the glass is half empty, and this is one of those occasions.
Thank you very much. On that note, we will conclude the discussion. I thank the Auditor General and his team for the information that they have provided.
Previous
Section 22 Report