Hansard

Good afternoon, ladies and gentlemen. The members of the Audit Committee are the first to receive copies of my report, which has just been laid before the Parliament. This public meeting of the Audit Committee is the proper place for my report to receive its first consideration. I would like to describe the most important features of the document—I hope that the committee will find this statement helpful. I will start by placing my report in context.

Members may have seen that the covers of the report are unusual—they fold out. That is to allow us to show graphical illustrations of what the new Scottish Parliament building at Holyrood will look like. It is important that we have these images in mind when we consider the report. The vision for Holyrood is not of a standard office block. It is a project to create a unique building, of quality and distinction, which will provide a suitable home for the Scottish Parliament for future generations. The successful completion of this new building would provide the Scottish Parliament with a distinctive home of historic significance. It is for others to say whether this unique building is required; my point is that the costs of such a structure do not compare with those of a standard office block.

It is unusual to conduct an audit examination of a project that is still running. Usually, auditors report after a project has been completed and the expenditure has been incurred. In the case of Holyrood, the project will not be completed for well over two years and, roughly speaking, some two thirds of the expenditure has yet to happen. It is important to distinguish between the actual costs that have been incurred so far and the estimated costs that will arise in the future. So far, the main concern of the Parliament has been the rise in the total estimated costs, most of which have yet to be incurred.

It is important that we understand what has happened so far, but it is equally important that we learn the lessons from this experience in order to ensure that the project is well managed for the remainder of its life. In my report, I make a series of recommendations for the future management of the project and I am pleased to say that those have generally been accepted.

Members will find those recommendations in the handout that I have provided, which is taken from page 9 of the report. Among other things, I recommend improvements in the risk analysis and financial reporting and I suggest that a cost plan for the project is urgently needed. I wish to emphasise, however, that there are still considerable risks, which must be carefully managed.

The architectural design is now settled and there is broad agreement between the various parties about the budget of £195 million, which includes an allowance for unforeseen contingencies. However, as I said, the progress of construction on site is still at an early stage. With more than two years' work remaining, the targets for construction are not guaranteed. In my report, I describe the risks and uncertainties that remain.

Without going into detail, I will say that there will continue to be risks associated with inflation in construction costs in the period between the original estimates prepared by cost consultants, who were working to a price base of March 1998, and the final costs in cash terms, which will not be known until the project is completed. Project management recognise that and have agreed to review the inflation risk with their cost consultants this autumn.

There will continue to be construction risk. Under the chosen procurement route, which is known as construction management, some risks can be transferred properly to contractors, but significant risks remain with the client. Those construction risks are also described in the report.

There is no doubt that Holyrood has been a complex and challenging project to manage. Very unusually, there was a change of client after two years, around the time that construction was starting on the site. Until June 1999, the client was the Secretary of State for Scotland. On 1 June 1999, the client became the Scottish Parliamentary Corporate Body.

The corporate body was charged with ensuring that the developing needs of the new Parliament, in its first year of business, were taken into account. Some of the changes requested by the corporate body—many at the behest of the Parliament—were very significant. Project management had the challenge of translating the new requirements into a specification for the design team. In turn, the design team faced the challenge of solving design problems against very tight deadlines. Finally, we have the construction manager, which is the firm co-ordinating the design and construction processes and the works contracts. Until this summer, the construction manager was required to manage the processes and contracts in the absence of an agreed and settled design.

Taken together, all those challenges were very significant indeed, and the various groups have achieved a great deal. The corporate body finally signed off the design in June 2000 and the new buildings are starting to take shape on the site. The plan is that the building should be ready for occupation in December 2002.

My report starts with an executive summary and recommendations and is then divided into three parts. There is a description of the project from its inception in 1997 until the point, during this summer, when a project group was established. As the broad sequence of events is well known and is laid out in some detail, I do not intend to say anything in particular about that part of the report. The second part is an analysis of the reasons for the increase in the cost estimates and the later delivery of the project. The final part is an examination of the management and oversight of the project. In the rest of my remarks, I will say something about the reasons for the cost increases, the later delivery of the project and how the project was managed.

The corporate body has reported that the estimated total project costs rose from £90 million in January 1998 to £195 million. The figure of £195 million remains the latest figure, which reflects the approved budget for the project. There are two main elements in the total project cost: the construction costs and the associated project costs, mainly involving consultants' fees, a contingency allowance, fitting-out costs and VAT. There will be related costs for landscaping and road realignment works, but those costs are not included in the corporate body's current estimate of £195 million. Exhibit 8 on page 25 of the report summarises the costs.

There has been some confusion in the public reporting of the various figures. It is wrong to compare the original construction costs of £50 million with the current total budget of £195 million. The construction cost estimates have increased from £50 million to £108 million. That is an increase of 116 per cent. Almost half that increase has been caused by a 47 per cent increase in the total area of the building, which now stands at 31,000 sq m.

Within any public building there is the usable space and the space that is required for stairs, lifts, engineering plant rooms and other needs. That is called the balance area. In all, we have seen an increase in the space of the building of some 10,000 sq m. About 3,000 sq m have been added to the usable space to reflect the wishes of the client. Some 6,300 sq m have been added to the balance areas and about 2,500 sq m have been accepted by the client as necessary, for example, to increase the size of the main entrance areas and to reduce the intensity of use in historic Queensberry House. Partly offsetting those increases is a reduction of 1,900 sq m in the area devoted to car parking.

The other main reason for the increased construction costs is that the design is much more complex than the notional ideas for the building at the feasibility stage, back in 1997. Exhibit C, which is inside the back cover of the report, shows an illustration from an architectural feasibility study that was completed in December 1997. That feasibility study was used in estimating the £50 million construction cost. As we can see from the other exhibits on the covers, the design team has developed an entirely different concept for a linked series of buildings of a much higher quality involving a significantly higher cost. We have estimated that there has been a 48 per cent increase in unit construction costs and in my report I identify the features of the current design that help to explain that.

Another reason for the increased construction costs is that the security of the building has required significant increases in the estimate. The final factor under construction costs is the relatively high cost of refurbishing Queensberry House. A final decision to retain Queensberry House was not taken until later in 1998 and the cost of refurbishing this building is now some £7 million—about £2 million more than the figure first expected by project management. That is because of the need for additional structural repairs, which were not detected in the initial surveys.

Finally, still under this heading of the reasons for the increased construction costs, I should mention that, in addition to the project costs for which the SPCB is directly responsible, there will be costs for landscaping and road works. All that will take place on land for which the SPCB is not responsible. On the outside back cover of the report, there is a site plan that includes an indication of those works. The Scottish Executive has given me an estimate that the costs could amount to £14 million, including contingencies of about £2 million. Those costs have been excluded from the project estimates because they will not be funded from the SPCB's direct resources; they are the responsibility of the Scottish Executive and funding will be taken forward as part of the current spending review.

As I mentioned, associated project costs have increased from £40 million to £87 million as a direct consequence of increased construction costs. At the outset there were site purchase costs of £5 million; those costs have not changed. Fees for designing and managing the construction project are now estimated at £26 million; there is a contingency allowance of £11 million; furniture and fitting-out is estimated at £17 million; and VAT will amount to £28 million, of which £4 million is likely to be recoverable from Customs and Excise.

A notable feature of the Holyrood project is slippage in the timetable for completion of the buildings. A diagram on page 26 of the report shows clearly how changes in the forecast programme for completing the building project have occurred. The most striking feature of the diagram is that, although the main architectural scheme design was initially intended to be completed in March 1999, it was not achieved until June 2000. However, construction work started on site pretty well when planned, namely in June 1999. My report looks at this issue in some detail.

In general terms, much of the extended time scale can be attributed to difficulties in achieving the approved design. Some of the delay arose from problems encountered by the architects in complying with the original demanding brief to an extremely tight timetable. Furthermore, the architects had the challenge of meeting unforeseen changes, which were requested by the client at various stages.

The third part of my report examines how far the project's organisation and the management and procurement processes complied with good practice. It also touches on some general questions about the governance of the project.

The structure for overseeing managing and delivering the Holyrood project is complex. However, that is not untypical of large public sector projects. Members might find it helpful to refer to exhibit 2 on page 5 of the report, which attempts to explain a quite complex structure. First, there is the client. Until June 1999, the client was the Secretary of State for Scotland; since June 1999, the client has been the SPCB.

Secondly, there is project management—a collective phrase referring to a number of people. Until June 1999, project management consisted of the principal accountable officer in the Scottish Office—or the permanent secretary—and the project team, which is the team of civil servants and private sector appointees on secondment to the team, and which is headed by the project sponsor, who now has the title of project director. After June 1999, project management changed. The clerk to the Parliament became principal accountable officer; however, the project team remained the same and moved over from the Scottish Office to the SPCB.

In addition, we must take into account the design team of architects, structural and service engineers and cost consultants. Finally, there are the construction manager and the works package contractors. Since the summer, there has been a new player on stage—the Holyrood progress group. As I said, project management have faced major challenges in managing in this complex environment, not least because of the change of client and the significant changes in the specification that came from that client as the project was running.

It was clearly important that project management consisted of people with the best experience and skills to do the job. The creation of the Holyrood project management team generally followed good practice and the team had a mix of relevant and appropriate skills. However, the particular type of contract, known as a construction management contract, was unusual in the public sector. Indeed, it was rather innovative; we have not been able to find a similar example. Although the contract offers distinct advantages in allowing the client to control—and indeed change—the specification as the project is being implemented, it has the possible disadvantage that more of the risk is left with the client rather than being passed to the contractor. As a result, I ask in my report whether the Holyrood project team always had available the best professional construction expertise to meet the demands of this large complex project. I wonder whether, at critical times, project management missed the presence of a senior experienced construction professional in at least one of the three key positions of project owner, project sponsor or project manager.

The appointments of all the consultants to the project were properly undertaken. However, it would have been helpful had some aspects of the appointments process been more systematic and better recorded. From the available information, there were clearly strengths in the selection procedures for the designer competition. There was a good response to the competition, with internationally renowned architects expressing an interest. I am satisfied that the appointment was based on merit and that the selection panel included suitable people, such as two eminent independent architects and the chief architect of the Scottish Office. The remit of the selection panel was appropriate and it is important to record that the panel was unanimous in its final choice of a design team.

Furthermore, there was a good response to the advertisement of the construction management appointment. However, I mention in my report that procedures for this appointment were not as systematic as they might have been. For example, the winning firm was the second highest tenderer; project management selected it after taking the quality of its bid into account and after obtaining significant financial adjustments to that bid. Although it was clearly right that project management should take quality as well as price into account, it would have been helpful if the supporting analysis had been better recorded.

I wish to emphasise that I do not consider that those procedural shortcomings on the client side have adversely affected what happened regarding the management of the project. I have no reservations in saying that the firm appointed as construction manager is eminently suitable for the job.

My report mentions the fees payable to consultants. All the appointments were made in 1998 and early in 1999 when the approved budget for construction costs was £50 million. At that time, the estimate of fees payable to the consultants was £10 million. As I explained, with the significant increase in the estimated construction costs, the estimated fees have also increased significantly, the latest figure being £26 million.

There is no doubt that consultants have engaged in significant additional work because of the various changes and problems encountered with the project. However, my report suggests that project management could have explored more carefully an alternative fee arrangement that might have shared the risks of cost increases between the client and the consultants. That would have provided an incentive to control costs.

I note in my report that other areas of project management did not match Treasury guidance on best practice. A particular concern is that there should have been formal procedures to control changes, based on a detailed cost plan that was agreed between all parties at an early stage. Another concern is that, largely as a result of the delay in achieving an approved design, a firm cost plan has still not been agreed between all the main parties; there must be risks associated with that.

Much of the concern that has been expressed about the project, both in the Parliament and by the public, arises from shortcomings in the arrangements for cost reporting. There was no arrangement requiring project management to provide full cost information to the client on a regular and systematic basis. Before June 1999, monitoring concentrated on core construction costs rather than on the full financial provisions that were necessary for fees, furniture and fittings, and VAT.

On some important occasions, project management did not report all the relevant construction cost estimates to the client. There were various reasons for that, but the main one was that the cost estimates were not acceptable to project management because they significantly exceeded the available budget. Therefore, project management concluded that they could not recommend them to the client and felt that the publication of those figures would reduce the incentive for consultants to get the costs down to the available budget. I acknowledge the importance of keeping pressure on the consultants, but the high estimates coming from the expert cost consultants made it more, not less, important that the client was informed, so that the client could make judgments at the highest level on the stewardship of the project.

A related issue is how risk was treated. For most of the life of the project, there was a general contingency allowance of 10 per cent of construction costs. However, an important shortcoming was that project management did not identify and quantify a separate allowance for the major risks that might affect the project. The separate identification and quantification of risk is an essential element in the Treasury guidance.

I will explain the recommended model. At the start of any project, no expenditure has been incurred. There is a base estimate of the cost, with no risk allowance. To the base estimate should be added an allowance for risk; that is calculated by a formal risk analysis, which considers all the factors that might come into play. As the Treasury guidance says, that is not a guess at contingency or a general slush fund. As the project develops and becomes more clearly defined, either the risks become a reality and are built into the base budget or they are avoided by good management. Any unidentified risks that arise as the project progresses should be managed by formal change control procedures, which feed into the base estimate and the working cost plan.

In the case of the Holyrood project, the cost reporting to the client should have been based on that approach and should have included an explicit analysis of possible risks and their financial implications. As I mentioned, for most of the project, that did not happen, and the costs that were reported to the client were the construction costs, including a general contingency allowance.

In February of this year, the corporate body asked John Spencely to undertake an independent investigation and report on the project. By April this year, Mr Spencely was referring to a construction cost estimate back in May 1999 of £89.2 million—a figure provided by the cost consultant. In response to a request from the First Minister for an explanation of why he was advised, back in May 1999, that the project construction cost was estimated at £62 million plus contingencies of £6 million, the permanent secretary reported that the figure of £62 million was project management's best judgment of the most likely construction costs, taking into account the progress achieved on the design so far.

Audit Scotland has looked at the figures and has confirmed to me that the estimate of construction costs made by project management was £16 million lower than the estimate of the expert cost consultant. That was because project management excluded certain risk allowances made by the cost consultant. The permanent secretary's report explains the reason behind the judgment taken by project management to exclude those risk allowances. I have included his report as an annexe to my own.

The specific risk items mentioned in the permanent secretary's report did not materialise or were overtaken by subsequent changes to the project. There remained, however, a significant risk factor throughout the design period. Reports by the cost consultant to project management between August 1999 and June 2000 have contained contingency allowances totalling between £8 million and £19 million. It is only with the acceptance of the scheme design for the project in June 2000 that we have seen a coming together of the views of project management, the cost consultants, the design team and the construction manager on the total expected costs of the project.

The latest position appears to be more satisfactory. The contingency sum is £11 million; that is consistent with the provision suggested by the expert cost consultants, whereas the earlier allowance was not. The design of the scheme is, of course, more advanced, which reduces the risk of cost increases as a result of uncertainty. However, I repeat that there are still significant risks associated with the project. I am suggesting that project management should look again at the overall cost provision in the light of those and other risks that may be identified. Project management must ensure that all the risks have been systematically identified and evaluated and that there is a proper, separate allowance for risk in current and future estimates.

My report concludes with a section about corporate governance of the Holyrood project. Although it is a short section, I believe that it is important.

An important element of good governance is the ability of the members of an organisation who are responsible for taking decisions about the direction of that organisation to take a sufficiently independent stance in relation to the permanent staff. The organisation in this case is the Scottish Parliamentary Corporate Body, which comprises the Presiding Officer of the Parliament and four members of the Parliament. The permanent staff are led by the clerk to the Parliament and include most members of the Holyrood project team.

A very important stage of the Holyrood project occurred when legal responsibility for it passed from the Scottish Office to the SPCB on the 1 June last year. At that moment, it would have been appropriate for those accepting responsibility for the project to have reviewed it with a degree of independence from the project team, in order to satisfy themselves about its status and health. There was no such review at that point and the independent assessment by Mr Spencely did not occur until the spring of this year.

I offer the suggestion that more should have been done to advise the members of the corporate body about their proper role in overseeing this project, about the proper role of the officials and about the unusual and innovative features of the contract management arrangements for this major project. With the benefit of hindsight, I also suggest that it might have been advisable to allocate the accountable officer responsibility for the Holyrood project to someone other than the clerk to the Parliament who, of course, carries an extremely heavy work load apart from his Holyrood responsibilities.

In June of this year, the SPCB established the Holyrood progress group. One of my recommendations was that the SPCB should consider whether there is any need for independent advice in future. A good feature of the progress group is that its membership includes an architect and a quantity surveyor, both of whom are independent of project management. The group is meeting regularly and is closely monitoring progress on the project. In addition, the SPCB has restructured project management to ensure that the right professional skills are available to manage the project within a tight budget and against tight deadlines.

In accordance with the request of the Audit Committee, I have identified and explained the reasons for the increase in the estimated costs of the Holyrood project and the slippage in the timetable, and I have examined the key features of the project management arrangements.

The facts in my report have been agreed with the accountable officers but the narrative and conclusions are my own. Although there is agreement on many of the conclusions, the accountable officers might take a different view on some of them and they might not share the emphasis that I have attached to some issues.

I am most grateful to everyone who has co-operated in providing information, often against tight deadlines, and I am also personally grateful to the members of the Audit Scotland team, who, over the summer, have worked with me in bringing this report to you.