1 Wednesday, 18 October 2017 2 (9.30 am) 3 CHAIR OF THE INQUIRY: Good morning. Yes, Mr Lake. 4 MR LAKE: My Lord, the first witness today is Mark Bourke. 5 MR MARK BOURKE (sworn) 6 CHAIR OF THE INQUIRY: Could I ask you to speak clearly into 7 the microphone and not too quickly, so the shorthand 8 writers can keep up with you. 9 A. Certainly. 10 CHAIR OF THE INQUIRY: If you listen to the question and 11 answer it as concisely as possible. If you want to add 12 an explanation, do so after the primary answer. 13 A. Thank you. 14 Examination by MR LAKE 15 MR LAKE: Mr Bourke, could you state your full name, please. 16 A. Mark Bourke. 17 Q. The Inquiry has your address. What's your present 18 occupation? 19 A. A consultant engineer. 20 Q. Could I ask you please to look at a document with me. 21 It's got reference TRI00000110. Now, Mr Bourke, that 22 will be shown on the screen slightly to your right, but 23 you also have a paper copy of it in front of you? 24 A. Yes. 25 Q. You will see that that is the statement that you gave to 1 1 the Inquiry? 2 A. That's correct. 3 Q. I think looking at the paper copy, you will see that you 4 signed it on the last page? 5 A. Yes, I did. 6 Q. Are you content that that be accepted as your evidence 7 to this Inquiry? 8 A. Yes, I am. 9 Q. What I would like to do is ask you some additional 10 questions about certain parts of it. So could we first 11 turn to page 9 of this statement. In fact it may be 12 easier to go back to page 8, just to get the start of 13 the particular answer. 14 Firstly, look at the question 11, which are the 15 first two numbered paragraphs 11 on the screen. It 16 says: 17 "How did the approach to and management of risk in 18 Tie compare with other organisations for which you have 19 worked? Was the assessment and management of risk 20 undertaken well at tie?" 21 I'll read through your answer and then I want to ask 22 you for clarification about certain of the subparagraphs 23 that are written. 24 What you say is: 25 "The approach to and management of risk in tie 2 1 compared well to the other organisations that I have 2 worked for and advised as I believe that there was 3 a strong senior level commitment to identify and 4 mitigate risk with the support of adviser expertise. 5 I believe that there are a number of key features that 6 make me conclude this (i) there was heightened awareness 7 of the potential risks affecting tram schemes and how 8 these could be mitigated at a strategic level through 9 the adoption of an innovative procurement strategy; (ii) 10 there was a conscious decision to develop a unified 11 project risk register with the support of colleagues and 12 advisers to ensure a common language and focus on risk; 13 (iii) a risk management policy was developed to 14 demonstrate Board level organisational commitment to the 15 management of risk." 16 If I just pause there, that's the first of the 17 subparagraphs I want to ask you about. Could you 18 explain what you mean by that, number (iii)? 19 A. Number (iii), essentially we developed an organisational 20 policy. At the time risk management was not an area 21 where all organisations actually created a policy and 22 had a policy. 23 We developed a policy document that demonstrated the 24 organisational responsibilities, their appetite, and our 25 approach to risk management. 3 1 Q. Simply the formulation of and adoption of that policy, 2 is that what provided a commitment to understanding 3 risk? 4 A. Absolutely. Like most organisational policies in 5 defining an approach, organisational commitment and 6 responsibility to address risk and to manage it for the 7 benefit of the public sector. Having a Board level 8 commitment to that policy showed that there was 9 a top-down commitment to address risk affecting both the 10 business and projects. 11 Q. We have seen in documents already looked at at the 12 Inquiry that initially the project risk register that 13 you referred to was considered by the Tram Project 14 Board. But after a while that was delegated, 15 consideration of that was delegated to a sub-committee. 16 Is that consistent with the maintenance of the 17 consideration of risk at top levels? 18 A. In terms of the terms of reference or responsibilities 19 of sub-committees, those, I imagine, would have to in 20 turn report up to the Board. So I had anticipated that 21 sub-committees would escalate matters such as risk 22 topics to board level, so the creation of 23 a sub-committee with a deeper consideration of risk 24 would not necessarily disconnect the Board from that 25 understanding. 4 1 Q. If we look at the next subparagraph in your statement, 2 number (iv): 3 "There was a broader focus to examine the areas that 4 had led to causes of increased project delivery and 5 those presented by Optimism Bias." 6 Could you explain that sentence to me? 7 A. In terms of broader focus, I'm emphasising there was 8 considerable effort made to understand the risks 9 affecting tram infrastructure construction. 10 This was effectively through the engagement of 11 advisers who had considerable experience in delivery of 12 infrastructure, and in addition a broader review of the 13 risks that present themselves through optimism bias. 14 Q. I was just puzzled in particular by what you mean by 15 things that had led to the increase of project delivery? 16 A. Increased project delivery, by that I mean increases in 17 cost and delays in schedule. 18 Q. And those presented by optimism bias, is that again the 19 increases in time and cost that you're referring to 20 there? 21 A. Yes. Yes, optimism bias focuses on cost programme and 22 benefits. 23 Q. I want to ask you some questions about optimism bias 24 towards the end of your evidence. 25 Look at the fifth subparagraph: 5 1 "tie invested in dedicated risk management resources 2 and sought to implement leading edge technology to help 3 manage risk through investment in Active Risk Manager." 4 Again, could you explain that? 5 A. Essentially we created a team embedded at project level 6 and corporate level to manage risk. That type of 7 commitment wasn't necessarily common within projects 8 inasmuch as the risk functions could have been provided 9 by supporting advisers and embedded in those 10 organisations. But tie's investment in creation of 11 a dedicated risk resource showed a serious commitment to 12 address risk, to manage risk for the benefit of the 13 public sector. 14 In terms of the technology we invested in, one of 15 the issues we had was the number of parties who were 16 involved in the delivery of the infrastructure, and that 17 was only going to increase as we moved into procurement 18 and delivery stages. 19 So the decision was made to look at a common 20 platform, in effect a piece of software that would allow 21 us to use a common software infrastructure to record and 22 manage our risks, and that was the software Active Risk 23 Manager. 24 Q. I want to return to the question of the use of third 25 parties for risk; first, if we could look at the final 6 1 subparagraph in this paragraph number (vi), where you say: 2 "The effect of our approach to risk management 3 included greater on-going attention to risk management 4 and greater dialogue resulting in risk forming part of 5 most discussions." 6 Could you explain that? 7 A. I think maybe picking on the last point in terms of 8 greater dialogue, one of the -- one of the features of 9 having a dedicated risk team and having a greater focus 10 on risk is ensuring that each meeting, discussion, there 11 is discussion regarding risk. So because of our 12 efforts, there was a greater dialogue, a greater 13 understanding and appreciation of risks, and effectively 14 that being step 1, understanding the risks that 15 potentially the project could be faced with. 16 In terms of the effect of our approach to risk 17 management and greater ongoing attention, having 18 a dedicated team who can focus on that and drive that 19 through the business and work with advisers and 20 suppliers allows us to increase our attention in that 21 area. 22 Q. Now, in response to my earlier questions, you have 23 talked about having a dedicated in-house resource, but 24 I think if you go on, if you look at paragraph 12, your 25 answer, at the foot of this page, I don't need to read 7 1 through that, I don't think, but what you are 2 considering there is the fact that at the initial 3 stages, the risk management was provided by the 4 designers under the SDS contract. 5 A. Yes, we specified our remit for the provision of risk 6 management services within the SDS remit. 7 Q. Now, is that common to use a designer to provide risk 8 management services? 9 A. It is. The breadth and detail that we went to in 10 thinking through the scope of risk management within the 11 designer was probably more detailed and developed than 12 had been done before. By that I mean we gave complete 13 clarity of our requirements, because we saw that the SDS 14 designer would have a role that would continue through 15 early design phases and then post-novation as part of 16 the -- design and build contractor would have a central 17 function, effectively managing their design-related 18 risks, but would have touch points to all the risks that 19 the entire project would be exposed to. 20 Q. If I could just ask you in particular about one of the 21 points you make in this paragraph we see on the screen, 22 four lines up from the foot of the page, beginning with 23 (iii): 24 "SDS would be managing design development risk 25 would be central to de-risking the project and would 8 1 have a direct influence on MUDFA, Tramco and Infraco." 2 Now, in what way was the design development risk to 3 be central to de-risking the project as a whole? 4 A. The design development was fundamental to enable the 5 procurement strategy that we were adopting. One, it 6 would allow the early utility diversion works through 7 MUDFA to be executed. Two, it would allow interface 8 with the operator, City of Edinburgh Council, to advance 9 the design to such a level that the key consents and 10 issues that would be considered during design 11 development could be effectively developed in advance of 12 a contractor, the main contractor, Infraco, coming on 13 board. 14 That was seen as a way of de-risking, rather than 15 employing a contractor and designer to effectively 16 commence at the same time, and then go through the 17 development stage. 18 It was also fundamentally to provide a product, 19 a design development that Infraco could effectively 20 price, and with that develop product, hopefully bring 21 less risk-related pricing from Infraco. 22 Q. The way you described it there, would it be fair to say, 23 trying to put it in a nutshell, that a lot of the risk 24 management and procurement strategy hinged on getting 25 the right design at the right time? 9 1 A. Absolutely. 2 Q. Was that something that, as far as you are aware, was 3 the subject of common understanding above the senior 4 management at tie? 5 A. I believe that there was no doubt in senior management 6 with regard to the procurement strategy which was 7 a strategy that had been developed and deployed as 8 a core methodology, a core approach to the business. 9 There was no ambiguity at all between understanding 10 procurement strategy and what we were trying to achieve 11 to effectively de-risk the works. 12 Q. Could I ask you then to look at page 6 of your 13 statement. If we highlight the whole of the paragraph 7 14 there. 15 You describe -- reading, I think it's eight lines 16 down, the sentence beginning: 17 "As the project further progressed, we sought to 18 develop the management of risk through SDS and 19 incorporated within their management services 20 a requirement to provide extensive risk management 21 services and deliverables which I had drafted, as 22 outlined in Section 4.2 of their Scope of Services." 23 That was the emphasis on requiring them to provide 24 not just the design, but the risk management service; is 25 that correct? 10 1 A. That's correct. 2 Q. You note then that one of the things they were to do was 3 to produce and maintain the project risk register, and 4 that within SDS, that was James Kimmance and 5 Paul Jobling were to undertake that? 6 A. Correct. 7 Q. Then critically you say: 8 "I also recall that we took a decision, following 9 poor performance of SDS in the execution of these 10 services, that we would develop this in-house with the 11 support of TSS, who provided Nina Cuckow to fulfil the 12 role of Tram Risk Manager." 13 A. That's correct. 14 Q. Is that essentially saying that the plan to use the 15 external resource and risk had in essence failed? 16 A. It hadn't been successful. We had to adopt an 17 alternative plan in order to ensure that we managed 18 risk, and that alternative plan was to use another one 19 of our advisers, TSS, effectively who would have been 20 a review, independent review of SDS's risk deliverables 21 to actually step in and take control of that. 22 Q. In the following sentence you note that: 23 "At a later point in time, we took the decision to 24 directly employ a dedicated Tram Risk Manager, 25 Mark Hamill." 11 1 A. Correct. 2 Q. Was it therefore at that stage really that the in-house 3 risk resource was developed when Mr Hamill joined tie? 4 A. Correct. 5 Q. Going back, though, to the SDS performance, what was it 6 about their performance that was poor and necessitated 7 their handing of the risk management to TSS? 8 A. They weren't producing deliverables on time, to the 9 quality we expected. They weren't engaging effectively 10 with the team and there was a lack of confidence that 11 they would have an ability to do that. 12 Q. When we are talking about deliverables, is that things 13 such as the risk register? 14 A. Yes. 15 Q. If we go, jumping forward then to page 42 of your 16 statement. 17 CHAIR OF THE INQUIRY: Before doing that, could I just ask, 18 you said that the plan was SDS would provide these 19 services and that TSS would be -- act as a reviewer -- 20 A. Correct. 21 CHAIR OF THE INQUIRY: -- of what they were doing. When you 22 decided to change to TSS as providing the services, who 23 acted as reviewer of TSS? 24 A. That's a good question, my Lord. The review that we 25 deployed on that was within the project team. So the 12 1 range of expertise that we assembled within in-house, 2 effectively, became the reviewers of those areas. 3 CHAIR OF THE INQUIRY: And then when Mr Hamill came to 4 provide the services, who reviewed his work? 5 A. With the support -- I believe with the support of TSS 6 advisers and the rest of our team, they provided that 7 review. These were, for clarity, my Lord, these were 8 ongoing day-to-day activities whereby each risks would 9 be discussed with risk owners and risk action owners. 10 In effect, the risk manager's role is to ensure that the 11 progress has been made not to necessarily identify that 12 all of the risks have been identified from the team or 13 the mitigations are adequate; but act as a one layer of 14 defence, but through broader review and the scrutiny 15 that we brought in through wider advisers, and even 16 additional review that was undertaken by 17 Transport Scotland's advisers at key stages. That 18 brought the comfort that the register that we were 19 preparing and managing was sufficient. 20 CHAIR OF THE INQUIRY: Thank you. 21 MR LAKE: Could we look at page 42 of your statement, 22 please. If you just highlight the second paragraph, 46, 23 the answer. Second one, please. 24 You note there: 25 "I requested Graham Nicol to prepare a paper 13 1 considering the options we had in relation to SDS's risk 2 management remit. I recall that this was prepared for 3 discussion with senior colleagues as we had reached 4 a point where the quality and omissions in their 5 deliverables was compromising the quality of risk 6 management on the project." 7 That's really referring to what you've already said 8 about the lack of deliverables on the part of SDS? 9 A. That's correct. 10 Q. How was it compromising the quality of risk management? 11 A. The period that this was taken was proceeding under -- 12 we weren't getting the traction and the progress that we 13 anticipated to have on these areas of risk. We were 14 very keen to progress risk in a number of areas and saw 15 the engagement of SDS as another way that we could step 16 up our risk management and make further improvements. 17 We anticipated, for instance, that SDS would, 18 through their design development, identify a wider set 19 of risks and be able to provide more detailed and 20 developed risk mitigations. 21 In effect, with not receiving deliverables that were 22 developed or well thought through, that effectively 23 compromised our ability to instruct mitigations to 24 proceed and us to effectively seize the opportunities 25 that we may have had. 14 1 Q. Once you had brought in TSS, Nina Cuckow in particular, 2 to what extent was it possible to recover from that? 3 A. That's a very good question. The ability to recover was 4 difficult. However, when we reviewed the options that 5 we had available to re-energise this risk review, we 6 considered options of blending SDS with TSS to have 7 different components of risk considered with each team; 8 and then concluded that the best option, moving forward, 9 which is in this paper that was put together, was to 10 progress with TSS. 11 That effectively brought team members from TSS who 12 were familiar with the project, who had had a previous 13 involvement, and were able to mobilise quickly and 14 effectively. However, the lost period of SDS not 15 performing effectively did have an impact. But it's 16 hard for me to recall exactly how to quantify that. 17 Q. Because it was a period, I think, where obviously the 18 efforts had been made to advance the various main 19 contracts that were going to be entered into for 20 delivery of the trams? 21 A. Yes. 22 Q. The early stages of the Infraco procurement had taken 23 place? 24 A. Yes. 25 Q. The MUDFA contract had been placed? 15 1 A. Yes. 2 Q. They were being done in an environment where it is now 3 acknowledged by you that the risk management was not 4 what it should have been? 5 A. Yes. 6 Q. Do you consider it had any effect on a proper 7 understanding of risks in the procurement processes that 8 were going on? 9 A. In relation to the risks in relation to procurement, 10 I believe those weren't affected in terms of the risks 11 of controlling and managing procurement. 12 However, the risks in relation to SDS's wider 13 performance in relation to their design development 14 undoubtedly would have a direct impact, and that was 15 effectively a day-to-day issue at this point in the 16 process, that the team were very conscious of and were 17 managing day-to-day. 18 Q. That is something I was going to ask you. The Inquiry 19 has already heard evidence that there were concerns 20 throughout this period as to the performance of SDS in 21 their provision of designs. So that was -- and that was 22 something, as you have said, at the heart of the 23 contract. So itself presented a substantial risk. 24 A. Yes. 25 Q. But the SDS, wearing their other hat of risk management, 16 1 were underperforming. Was that compromising the ability 2 properly to comprehend and address the design risk? 3 A. The two -- the two issues would definitely compound each 4 other. The fact that SDS were not addressing their risk 5 management remit and were not producing the deliverables 6 necessary to effectively inform these wider contracts 7 had a compounding effect. 8 Q. Is that not something that's going to arise in any 9 situation where you've got the person -- the entity 10 that's carrying out the risk management is also 11 providing a separate service to a project, that the risk 12 involved in that second service is at danger of not 13 being properly reviewed? 14 A. Yes, there is a danger of that. In effect we wanted SDS 15 to -- because they were executing such a critical 16 element of design development and employing considerable 17 amount of staff, that they would be directly in touch 18 with the issues at a very detailed level, a very 19 granular level, and therefore they would be best placed, 20 rather than an external organisation, to identify risks 21 and come up with suitable mitigation plans. 22 Q. Because it might be thought that having someone else, 23 for example, considering the risk out of the design, 24 would give a greater objectivity, perhaps a greater 25 scrutiny or oversight of the issues that -- the risk 17 1 issues that arose in relation to design? 2 A. Absolutely. TSS, when they came on board, effectively 3 were providing that degree of support. 4 Q. To tie? 5 A. Through tie. 6 Q. So would it be fair to say that there was a better 7 understanding of the design risks once TSS assumed the 8 role of risk managers? 9 A. Definitely. I think the initial understanding that we 10 had, however, was not coming from a position where we 11 were oblivious to the key design risks that affected 12 projects. I would emphasise, we had spent considerable 13 efforts to understand the risks affecting wider tram 14 infrastructure projects, and considering the local 15 circumstances that we had, identified a number of key 16 mitigations to be addressed during the design 17 development. 18 Q. You mentioned in an earlier question I asked that -- in 19 response to an earlier question I asked, that it's not 20 uncommon for designers to undertake the risk management 21 exercise of projects. How is the lack of objectivity in 22 the assessment of design risk addressed in those 23 projects? 24 A. It's very common, in fact it's standard practice, that 25 designers maintain a risk register. This is 18 1 particularly in relation to their key decisions that 2 they're making. They, under good practice, should be 3 informing clients regarding the risks in their design 4 development and in their solutions. 5 In terms of objectivity, there's always some degree 6 of objectivity in that that role is allocated typically 7 to maybe one person within a design team to extract the 8 key risks from the design team and summarise that. 9 The objectivity or the benefit of having an external 10 look at that issue, an experienced external, is that 11 they can quite quickly focus on what the key issues are, 12 rather than go for the volume of issues to be addressed. 13 So there are advantages and disadvantages of 14 being -- having an embedded element and an overview 15 provided by an external. 16 Q. I want to ask you some questions now about how risk was 17 managed when it was being dealt with internally within 18 tie. Could we look at page 52 of your statement, 19 please. 20 I do this just to highlight the question first of 21 all, at the foot of the page. See what we're talking 22 about. The question is: 23 "What oversight was there to ensure that important 24 risks were not overlooked? That is if there was an 25 omission from the risk register because a risk was not 19 1 recognised, what measures were in place to detect and 2 correct it?" 3 If we can go over the page to see your answer, 4 highlight answer 58. I will read the whole of it just 5 to provide the context: 6 "I believe that the approach that we took was geared 7 to create a culture of active participation in the 8 identification and management of risk. We aspired [to] 9 implement risk management best practice and apply an 10 innovative procurement strategy. We spent considerable 11 effort to learn the lessons of previous projects that 12 led to a new procurement strategy that sought to de-risk 13 the overall project to the public sector. In 14 particular, we focused on those risks that were shown to 15 have contributed to Optimism Bias. Our governance 16 oversight included participation from TS, CEC and 17 Partnerships UK. We regularly sought the inputs of 18 legal, financial and technical consultancy to advise us 19 on the management of risk. Risk formed a key agenda 20 item in most meetings. We briefed stakeholders 21 including CEC, TS, and SG on our management plans, 22 primary risks, and mitigation planning and proposed risk 23 management. In addition, we sought independent scrutiny 24 to verify our approach. Dialogue within the project 25 team, including workshops, and our advisers considered 20 1 that management activities to ensure risk awareness risk 2 management. If a new risk was identified it would be 3 added to the risk register, appraised for significance, 4 assigned a risk owner providing oversight to that work 5 stream and the mitigation plans developed in conjunction 6 with the risk owner including the allocation of 7 responsibilities for implementing the mitigation plans. 8 Once we invested in Active Risk Manager, to create 9 a further integrated platform, this led to further 10 detailing of risk and planning." 11 Now, particularly in the second half of that 12 paragraph there, you set out what's done when a new risk 13 is identified. But I'm not sure you have really quite 14 addressed the question: what measures were in place to 15 ensure that risks were not overlooked? To identify not 16 what was done with the risk once it was in the register, 17 but the other risk that something never got into the 18 register? 19 A. Yes. In effect, the engagement with the contributors to 20 the project was that way of identifying new risks. We 21 had representatives in each one of our service providers 22 or advisers that we engage with to understand what they 23 perceive from their work streams to be the key risks, 24 and this is a -- took the form of workshops, email 25 requests, meeting discussions, that effectively generate 21 1 a series of potential risk events that need to be 2 considered in risk. 3 In terms of my view, we had a very good 4 understanding because of the strength of our team about 5 the types of risks that affected our projects. We 6 uniquely had an operator on board from a very early 7 stage, and that allowed us to really consider some of 8 the downstream issues much earlier that could have 9 tripped the project up. And through that broad 10 participation, which actually was further enabled 11 through using Active Risk Manager that allowed us to 12 effectively engage a broader set of the organisation 13 than just necessarily leaders, allowed us to identify 14 those risks. 15 But I would emphasise that the amount of study and 16 research that was done to identify lessons from tram 17 schemes, broader infrastructure, large scale project 18 developments, was considerable, to try and capture. 19 So I had complete confidence at the time that we 20 were scanning and spotting the risks in an appropriate 21 manner. 22 Q. We then turn to the question of identifying risks to 23 assessing them when they go into the risk register. If 24 we could turn to page 4 of your statement in this 25 regard. 22 1 If you could highlight paragraph 6. You describe 2 what was done there to assess risks. You say: 3 "These included qualitative assessments of (i) the 4 timing when that risk could feasibly occur during the 5 project lifecycle, including procurement, construction, 6 operation stages; (ii) potential impact areas, including 7 capex, programme, opex, quality, approvability; (iii) 8 potential Optimistic Bias risk impact areas; and (iv) 9 likelihood of occurrence and level of impact to assess 10 significance prior to and following mitigation." 11 I think you go on to say that a lot of those 12 assessments were carried out on -- one-to-one 13 discussions carried out with the individual people 14 involved in the project. 15 Now, in what context did those one-to-one 16 discussions take place? Would it have been an informal 17 discussion or a formalised list of questions and 18 directed discussion? 19 A. It was effectively a mix. To illustrate, one approach 20 would be to review the risks that the risk owner had 21 been allocated and test whether or not further 22 mitigations could be developed. 23 Also, to test their progress in implementing the 24 mitigation plans that they had. And then as an exercise 25 to identify anything new that has emerged during the 23 1 period since the last discussion. 2 The overall aim to identify the key risks that were 3 affecting the project. And having these one-to-ones, 4 you get an added advantage of a very open and engaged 5 style. Obviously with different individuals you adopt 6 different approaches, but generally engagement with the 7 team members right from the start in terms of legal, 8 technical, financial advisers, and within the team 9 allows you to pick up on the key specialist area risks 10 and the broader risk to the project. 11 Q. If you are speaking to a team member in relation to 12 a particular risk which arises within their remit, would 13 that team member be the one responsible for identifying 14 the possible consequences and the probability or 15 likelihood of any of those consequences occurring? 16 A. That would be -- could be that individual or it could be 17 done separately at a workshop. So, for instance, the 18 initial appraisal, effectively a review of the risk 19 significance, which looks at probability and its broad 20 level of impact, that would be done with the individual 21 to understand really the level of significance and 22 importance. 23 In that way it allows us to effectively consider the 24 effectiveness of our mitigation plans, because we 25 understand the significance of the risk originally, and 24 1 then, with considering with that individual the 2 mitigation plans, the significance after that. 3 Q. This is one of the things I was going to address. If 4 the individual in question who has -- it's within their 5 remit that the risk arises, they are also responsible 6 for identifying ways of mitigating that risk, but not 7 only do they have to identify the mitigation means, they 8 also assess the effectiveness of the mitigation? 9 A. Correct. 10 Q. Now, what arrangements are in place to verify or 11 cross-check that their assessment of the mitigative 12 effect is a good one? 13 A. Well, the measures that are in place are effectively 14 review and, you know, at project level, that can be with 15 project managers, project director, commercial 16 directors, et cetera, looking at the solutions that are 17 being deployed. Bear in mind that the mitigation plans 18 that are there, they require client instruction to 19 proceed. So there is an inherent review process that 20 goes on to decide where to proceed. For instance, 21 uncertainties regarding ground conditions. If 22 mitigation plans may include elements such as additional 23 surveys, ground surveys, those works would then need to 24 be scoped in detail elsewhere, and a decision to proceed 25 made by effectively ourselves. 25 1 Q. But taking that particular example because it's quite 2 a good one, if there was a risk of ground conditions, 3 a view can be taken as how likely it is that you will 4 encounter unexpected ground conditions; that's stage 1, 5 isn't it? 6 A. Yes. 7 Q. Then you have to think how to mitigate that risk, which 8 might be to carry out some sort of investigations to 9 determine what ground conditions are. That would be 10 stage 2? 11 A. Yes. 12 Q. And stage 3 would be assessment of how effective those 13 surveys would be in mitigating stage 1? 14 A. Correct. 15 Q. What if it starts to turn out that despite putting those 16 surveys in place, unexpected ground conditions are still 17 being encountered, so that it appears that it hasn't 18 been an effective mitigation of the initial risk? 19 A. This is a constant worry with -- in risk management, 20 that effectively is one, your mitigation planning 21 effective enough, and two, is your residual risk still 22 going to be significant? Are you -- and hence why the 23 client decision-making has to be there to see that there 24 is value in implementing some of the mitigation, or they 25 are effective. 26 1 In terms of that residual risk that even after, for 2 instance, conducting surveys, there is still a residual 3 risk. There are some risks that, quite frankly, you 4 can't eliminate. You can't avoid. And those 5 effectively, you are, for instance, in terms of ground, 6 seeking to understand the problem more, to hopefully 7 reduce the levels of contingency or uncertainty in the 8 problem. But then pass that information to, for 9 instance, the -- a contractor, the Infraco, to develop 10 his solution, in the knowledge that he may have to 11 supplement that with wider mitigation plans of his own. 12 Q. Even after risk is mitigated, there's going to remain 13 a residual risk or is likely to remain a residual risk? 14 A. Correct. 15 Q. That residual risk still has to be assessed and is 16 likely to be valued for the project? 17 A. Correct. 18 Q. And that value goes into -- in the tram project went 19 into a Quantified Risk Assessment? 20 A. Yes, that helps to inform the -- the quantitative risk 21 analysis that is undertaken effectively looks at all the 22 risks that are in the risk register, and then, through 23 an assessment of typically three points of the lowest it 24 could be, the highest it could be, and the most likely, 25 in assembling that across the risk register, can then 27 1 perform an analysis to quantify that. 2 Q. But who makes that assessment when looking at the 3 consequence of what the least serious, the most serious 4 and the most likely financial cost of that risk would -- 5 A. Typically, that's QS support. So a quantitative -- 6 a quantity surveyor who has experience in cost of -- 7 costing projects. 8 Q. And that would be an in-house quantity surveyor working 9 with the person who first identified the risk? 10 A. It could be a QS from in-house part of the adviser team, 11 part of TSS, part of SDS who could perform these 12 assessments. 13 Q. What I would like to do is go back to the example we 14 were talking about, ground conditions and surveys, and 15 look at the situation where surveys have been identified 16 as mitigation, but the problem still manifests itself, 17 so the mitigation hasn't worked. That would tend to 18 indicate that the figure that has been put on, 19 quantitative figure put on that risk is likely to need 20 to be revised because the mitigation isn't working as 21 well as expected; is that correct? 22 A. If the individuals looking at the specific risk were to 23 assess that and find, for instance, in sight of the 24 survey results that there is a greater degree of 25 uncertainty, they may find varied conditions, for 28 1 instance. They may still think that the risk is 2 present, and therefore the assessment they would 3 undertake to quantify the potential cost impact of that 4 would be in the knowledge of that; rather than 5 a revision of contingency, it would be an assessment in 6 the knowledge of the information at the time. 7 Q. You are looking at a situation where, for example, if 8 the risk does identify that there was going to be 9 problem ground conditions in particular areas. What I'm 10 thinking of is a situation where the surveys are to all 11 intents and purposes clear, or they find some problems, 12 then works take place and more problems are found. The 13 surveys have failed to find things. 14 Now, in that situation, do you have to conclude that 15 the original mitigation, getting surveys, hasn't been 16 effective to address the risk of unexpected ground 17 conditions, and go back and re-assess that risk and 18 possibly increase both the probability and the likely 19 cost impact of it? 20 A. Correct. 21 Q. When is that done? How is it done? 22 A. That effectively would be done through the review with 23 the individual risk owner who, in knowledge of the 24 survey results, would advise and update through Active 25 Risk Manager or a standalone risk register, and revise 29 1 the residual risk significance and update the cost 2 allowances, the three points estimate around that risk. 3 Q. So it's possible in that situation to have the 4 assessment -- the residual risk as assessed increasing 5 over time if further problems are coming to light? 6 A. It's possible. 7 Q. Also, dealing with the situation where the risk is 8 something that arises out of one of the contracting 9 parties not adequately performing their services, if 10 there was an increasing sign that they are not 11 performing properly, that could increase the probability 12 and the consequences of risk, such that residual risk 13 assessment would increase over time? 14 A. Correct. 15 Q. You've mentioned the review of risk, risk reviews, 16 rather than it being by the individual, it going to 17 meetings where these are considered and reviewed. How 18 do you identify which risks go to that sort of meeting? 19 A. It depends. If you're reviewing an individual's set of 20 risks, we would, for instance, if we were considering 21 the risks in relation to one specific area such as 22 utilities, for instance, we would gather together 23 those -- that individual's leading on specific risk and 24 go through the risks that they owned, and review whether 25 or not the risk is still valid, the effectiveness of the 30 1 mitigation, and where they are in progressing that 2 mitigation. 3 The review process is effectively to allow us to 4 update the risk register, but also allow us to 5 understand that we are making progress to implement the 6 mitigation. That being a key issue that we sought to 7 report internally in terms of: are we identifying new 8 risks? Are we closing risks if at all possible? And 9 how are we progressing with regard to our mitigation 10 plans? 11 Q. And if the mitigation plans are not progressing, does 12 that also require a re-assessment of the residual risk 13 on the basis that it might not be reduced in the way 14 that it was intended? 15 A. That's correct. 16 Q. What if the mitigation is proceeding but it's not having 17 all the desired effects? It's not being effective as 18 mitigation? 19 A. Typically the mitigation would be revised. It would 20 drop in to become a day-to-day management issue that the 21 individuals who are maybe managing that individual are 22 informed or the case that I'm thinking about, the -- for 23 instance, in relation to SDS, alternative fall back 24 measures are thought through and then implemented. 25 Q. Could we look at page 11 of your statement, just for 31 1 a moment. 2 Paragraph 16, the two paragraphs 16. 3 I had asked the question there about: 4 "The reports in the Tram Project Board papers 5 routinely note that risks have been reviewed. What did 6 that entail? Who did it and what records were kept of 7 this? What was the purpose of the exercise? Was it to 8 identify new mitigation measures or was it intended to 9 quantify the risk presented to the project and make 10 adjustments to the budget estimates? There is reference 11 in the report with the June papers to the purpose being 12 to ensure that the QRA output was as accurate as 13 reasonably possible. What did this involve?" 14 In your answer you have noted that the particular 15 example that was offered there was dating from 16 June 2009, by which you had left tie? 17 A. Yes, some two years after I had left. 18 Q. But all these things that I'm discussing there, did they 19 go on within tie while you were still there? 20 A. In relation to -- in relation to reporting of risk, that 21 proceeded in terms of reporting on QRA analysis, that 22 was undertaken. 23 However, I think by this stage, it appears as though 24 there is a process under way whereby QRA and contingency 25 are being assessed and regularly reported on. At this 32 1 stage that I was involved, we were formulating the 2 approach to ring-fence contingency to contracts and 3 manage the drawdown of contingency. 4 So apologies, I'm not clear how it actually panned 5 out after I left. 6 Q. I was more interested in what had been done while you 7 were there, and were you aware what was involved in the 8 review of risk by the Tram Project Board while you were 9 still with the company? 10 A. Well, as I say, we regularly reported on the risks. 11 Effectively we presented a primary risk register, which 12 was effectively a top slice of the most important 13 showstopper or critical risks to the board, to make them 14 aware of the key issues that we had and those included 15 some of the things that we've discussed earlier. 16 Q. You mentioned contingency there. We've also earlier 17 mentioned optimism bias and we've been discussing risk. 18 What is the relationship between risk, contingency 19 and optimism bias? 20 A. Right. Well, optimism bias, as you appreciate, is 21 a tendency for projects to underestimate their outturn 22 costs, and therefore is designed to effect a safety 23 valve or a headspace that sits above project costs. 24 Q. Just if I can interrupt there, what happens is an 25 assessment is done about the possible effect of optimism 33 1 bias; the result will be, as you say, that a sum of 2 money is identified as a sort of proxy for the 3 propensity to underestimate, and that is added back to 4 the estimated project cost as, in a sense, a correction 5 for that bias? 6 A. Correct. Correct. It's a forced correction that 7 through studies the Treasury have done, have identified 8 this feature and the level of allowance that should be 9 incorporated. 10 However, during the development stages of projects, 11 as projects progress, optimism bias, through risk 12 management, through greater understanding of a broader 13 set of risks, effectively has been shown to reduce. 14 So the -- essentially the approach that we took was 15 to understand a much broader set of risks that the 16 project could be exposed to, because optimism bias 17 identified a number of key features and lessons for 18 areas that had specifically driven cost overrun and 19 programme delay. 20 In terms of risk, risk obviously sits above both the 21 base costs and any uncertainty that's in the base costs. 22 This in turn is calculated at various stages during the 23 development of the project, and like optimism bias, it 24 doesn't necessarily stay the same. Both base costs vary 25 during the development of the project as more certainty 34 1 is introduced, and in effect you're eroding optimism 2 bias, which captures those broader unknowns or market 3 variables that you're not clear on. 4 Q. The risks in question, as you discussed the process of 5 the risk registers, it's trying to identify things that 6 might happen, or I suppose that might not happen, that 7 will have an effect on the cost of the project for or 8 the time within which it has to be completed? 9 A. Yes. Yes. 10 Q. They are known to the extent that a particular event or 11 non-event can be named, and the consequences of it, 12 financial or time consequence assessed for it? 13 A. Correct. 14 Q. With optimism bias, it's really, as you've explained, 15 trying to consider the way that people assess problems 16 or risks that arise? 17 A. Yes. 18 Q. And how they might underestimate the likelihood of it 19 happening or underestimate the consequence when it does 20 tend to happen? 21 A. Correct. 22 Q. Now, the way I have described them, I have done it quite 23 deliberately, they don't seem to overlap. They're very 24 much complementary? 25 A. In some sense I think they do overlap, because 35 1 effectively your optimism bias, for instance, if we pick 2 two areas that have driven optimism bias, one, the lack 3 of robustness of the Business Case, two, lack of clarity 4 around the scope of the client's requirements. 5 If you tackle those two issues as known risks, and 6 embed them within your planning, you're effectively -- 7 you are managing those areas that other projects, 8 through maybe the inadequacies of their Business Case 9 process, which might only involve writing a few pages to 10 get justification to invest, versus the degree of 11 scrutiny that projects such as the tram receive in going 12 through a large gestation period, a parliamentary 13 scrutiny, further design development that evolves over 14 years, that effectively have a longer term ambition to 15 address those compared to other sectors that don't have 16 that benefit. 17 Q. But in the question of -- you say a risk and the two 18 examples you have given are identified and addressed, 19 does the optimism bias not come into the addressing of 20 those in the sense that there's a tendency to be unduly 21 optimistic as to the extent to which you are able to 22 address the risk? Or the consequence of your attempting 23 to address it? 24 A. There is a potential for that, and guidance has 25 highlighted that particular concern, and it's also 36 1 highlighted the opposite issue, that if projects were to 2 proceed with large optimism bias included, that you have 3 an opposite behaviour whereby one, the marketplace knows 4 you've got more money and there's complacency to not 5 manage risk. 6 Q. I want to come back to the market knowledge of this. We 7 have talked about optimism bias and risk. The third 8 element, contingency, where does that fit in relation to 9 the other two? 10 A. Sorry, in terms of? 11 Q. Where does it fit in relation to risk and optimism bias? 12 A. In terms of contingency, contingency is obviously -- 13 well, I see it as very distinct. It's effectively the 14 interpretation of your QRA analysis that appraises your 15 risk in making a decision on the level of contingency 16 you want to proceed with in the project. 17 Now, the processes that we employed were to 18 effectively use QRA to analyse a range of conditions 19 that could occur for the risks that we understood, and 20 then in selecting a level of contingency to be 21 incorporated. 22 Q. So you're using contingency then, as I understand it, as 23 the figure that is to be derived from the quantitative 24 risk analysis? 25 A. Correct. 37 1 Q. Essentially it's the risk allowance? 2 A. Yes, it is. It's a risk allowance to be effectively 3 controlled by a project to deal with those risks that 4 have been identified. 5 Q. And in terms of -- you say in choosing a figure, if 6 you've done a quantitative risk analysis, my 7 understanding is it will produce a distribution curve 8 where you can choose what degree of probability you 9 want, and the higher the probability, the higher the 10 risk will be? 11 A. Correct. 12 Q. The higher the sum will be? 13 A. The analysis allows us to select based on our appetite 14 to the risk, level of risk that we are prepared to bear. 15 Q. I want to turn now to look at optimism bias, and I think 16 you have explained your statement that while you were 17 there, optimism bias was a calculation carried out. You 18 assessed it throughout your time on the project. 19 Did you expect that there would be a continuation to 20 the assessment of optimism bias once you left the 21 project in 2007? 22 A. As we moved to the Draft Final Business Case stage, we 23 had shown that the allowance of a higher probability of 24 risk occurring, ie a belief that we had identified 25 a broad range of risks, much broader than we initially 38 1 had at the very, very early stages, and then had 2 progressed to analyse these, that if we were to consider 3 a very extreme set of circumstances where all the risks 4 were to occur, and have a very high impact, that that 5 would be a far more onerous case than bolting on an 6 optimism bias contingency. 7 In addition to that, the guidance shows that the 8 allowance for optimism bias is primarily to be 9 incorporated at those early stages of the project, with 10 the aim to mitigate risks down to a residual level at 11 the point where you're appointing the Infraco contracts. 12 So the overall aim is to understand a broader set of 13 risks, analyse them much more effectively and in 14 a detailed fashion through a QRA, rather than in a very 15 simplistic inclusion of an allowance. 16 However, I would make the point that there's nothing 17 to stop funders or others to incorporate optimism bias 18 as reserve contingencies, in the knowledge of approving 19 an allowance with a contingency. 20 Q. You dealt with optimism bias in the early stages there, 21 but you also referred to it towards the end. We will 22 deal with it in reverse order, just looking at the end 23 first, where you said you decided to use a risk figure 24 to meet a higher level of probability, and you 25 considered that was more onerous than optimism bias. 39 1 When you say more onerous, do you mean it generated 2 a higher number, a greater allowance had to be made? 3 A. Yes. If you compare the residual allowances that the 4 guidance stipulates for, capex and programme schedule -- 5 Q. That optimism bias allowance -- 6 A. That optimism bias allowance versus the differences 7 between, for instance, a P50 and a P90, you'll find that 8 they're very comparable. 9 Q. They may be comparable in number, but they're not 10 assessing really the same underlying thing, are they? 11 A. In theory optimism bias is still catering for a residual 12 amount or a degree of unknown risks that may emerge. So 13 the potential is that there could be a broader set of 14 risks that could emerge. 15 However, the confidence we had at the time because 16 of the degree of huge investment we'd made to understand 17 and engage with a full supply chain, all team members to 18 understand and manage these risks, we were in belief 19 that this was the best approach. 20 Q. In terms of the initial assessment, if we could look 21 back to your statement, please, at page 33, if we could 22 highlight the central third of that page. Stop there. 23 We see just above the halfway line on the screen, 24 there's a sentence that begins: 25 "I was involved with determining the OB reference 40 1 class to be used and determined that it should be 2 classed as a "Standard Civil Engineering" as this seemed 3 the most appropriate for the scope of works being 4 contemplated. I recall that this further reinforced by 5 the example projects used to categorise a "Non Standard 6 Civil Engineering" Project that included complex, first 7 of kind unique, or incorporated high risk construction 8 elements including tunnelling works." 9 In that regard, as to which reference class you 10 chose, could we also look at page 49. The start of 11 paragraph 54, if we highlight that. Again, there you 12 note: 13 "The tram project was appraised as a "standard civil 14 engineering project", as the scope and characteristics of 15 the proposed construction works comprising roadworks, 16 utility diversion and tram infrastructure that did not 17 have unduly heavy risk elements." 18 Then you go on to talk about EARL which had 19 a different assessment as it had a tunnel in it. 20 Now, this project was, was it not, unusual in 21 putting a tram through a historic and congested city 22 centre? 23 A. Not unusual in terms of previous other light rail 24 projects that had been done in the UK or 25 internationally. It didn't have particularly heavy risk 41 1 elements in terms of non-standard features that Treasury 2 would point you towards adopting a non-standard 3 approach. 4 Q. But did you look at the other sort of projects that were 5 included within the standard civil engineering projects 6 and assess to what extent the tram might be different 7 from those? 8 A. Yes. I mean, the standard civil engineering projects 9 that were in there, we spent considerable time reviewing 10 this with -- in detailed discussions, as I recall, with 11 Transport Scotland, to basically understand and help 12 inform them regarding our decision to classify it as 13 that project. 14 That classification took place at a very early stage 15 of the process. It literally was in a matter of months 16 of me when I first started with tie. This was one of my 17 first tasks to basically examine and understand. So 18 that these assessments could be incorporated within the 19 STAG appraisal and considered within the early outline 20 Business Case. 21 Q. Who within Transport Scotland did you discuss this with? 22 A. I recall speaking with Paul McCartney and Damian Sharp. 23 I can't recall names of others. 24 Q. Were they content that it be classed as standard rather 25 than non-standard? 42 1 A. There was never any inference that non-standard was the 2 way to go with tram. 3 Q. Because obviously the guidance says there should be 4 a larger allowance for optimism bias at the outset for 5 a non-standard project? 6 A. I think one of the sort of main reinforcing points was 7 that when we looked at the other UK schemes that had 8 been developed, the outturn costs of those projects 9 relative were something like 25 per cent increase, and 10 therefore the idea of adopting a non-standard that we 11 built in a far, far greater allowance than that just 12 seemed to be an allowance that would just compromise the 13 deliverability of the system. 14 Q. But dealing with those other projects, the 25 per cent, 15 that was an increase of 25 per cent over the final 16 budget, what they thought they were going to cost. 17 A. Yes. 18 Q. Did you have estimates for those other projects of what 19 the final cost had been over their very initial 20 estimates to see to what extent they might have been 21 optimistic from the very outset? 22 A. I can't recall, but -- 23 Q. That's when the assessment would have to be made. 24 That's when the initial optimism bias was applied? 25 A. I would disagree, because in effect the outturn costs of 43 1 those projects are documented and therefore their cost 2 per kilometre, which can be corrected and adjusted for 3 time, location, et cetera, which we did as an exercise, 4 allowed us to do that. And therefore when we were able 5 to convert our basic costs per kilometre construction 6 relative directly to other UK schemes, it effectively 7 creates a reference class that's far more accurate than 8 the fact that we didn't know what the original pricing 9 way back was, when they first thought of an outline 10 Business Case for those schemes. 11 Q. Looking at other schemes and benchmarking against their 12 costs is one means of estimating what the actual costs 13 of the project will be. And that was something that was 14 done by tie; is that correct? 15 A. Yes. We employed advisers who helped us to develop the 16 cost models for these projects, for our project, who had 17 been involved with some of these previous projects. 18 Q. Indeed, but the point is that if you're once using the 19 other projects to arrive at your initial cost estimate, 20 to go further and try and adopt the same projects to 21 identify your optimism bias figures, or to dictate the 22 optimism bias class is in a sense over taking into 23 account or double counting those other projects? 24 A. I disagree. I think the project is a standard civil 25 engineering project. It's not got, you know, 44 1 unconventional features. It's not the first time 2 a project like this had been done in the UK. It doesn't 3 have high risk construction elements in terms of -- such 4 as tunnelling that have been shown to be drivers to 5 large cost increases in projects. 6 Q. If the Inquiry hears evidence from people involved that 7 it was a complex and challenging project, I take it you 8 wouldn't disagree with that? 9 A. It was complex in terms of yes, it's a large scale 10 complex project, but it's not the first of its kind. 11 It's just large and complex. 12 Q. But in meeting the various complexities and challenges, 13 and having to bring to mind the experienced minds of the 14 various engineers, it's in those minds that the optimism 15 bias may reside. 16 You have said a minute ago that taking a larger 17 figure would affect the deliverability of the project. 18 Obviously if you take a larger figure at the outset, 19 that is going to inflate the initial estimates, and 20 there's a danger simply that the funders, sponsors will 21 take fright when they see those initial estimates? 22 A. I don't know if it's take fright. I think it's the 23 reality in that if we were to proceed with all projects 24 and just bolt in the -- a large allowance, I don't think 25 funders would appreciate that. 45 1 Q. The concept or understanding of optimism bias, I think 2 really started to emerge in about 2002/2003; is that 3 your understanding? 4 A. In terms of the concept, the early concept? 5 Q. Yes. 6 A. Yes, I mean, the concept came earlier, but it was 7 crystallised by Treasury in their analysis in 2002/2003. 8 Q. As a psychological phenomenon it had been understood for 9 some time, but it started to be applied to construction 10 projects and costs in 2002/2003 by the Treasury? 11 A. Correct, and they mandated that through the Green Book. 12 Q. But it was about that time that the initial budgets for 13 the tramline in Edinburgh were being fixed. Were you 14 aware of that? 15 A. In terms of fix, do you mean -- 16 Q. The initial assessments of how much this was going to 17 cost in Edinburgh. 18 A. Well, it's part of our development of the outline 19 Business Case. The first Business Case I was involved 20 with, we included our assessments of optimism bias 21 within that submission. And as I say, the wider 22 documentation such as the STAG appraisal that looks at 23 the economic argument or economic case for the scheme 24 also incorporated optimism bias. 25 Q. As you said, both in terms of trying to show that 46 1 a project stacks up on the economic appraisal, the STAG 2 analysis, or both the funders be sure that they are 3 going to be able to finance it, there is a pressure to 4 reduce optimism bias? 5 A. There is -- there is reports published about that 6 tendency that there could be an in-house driver to 7 reduce optimism bias, but -- 8 Q. Because a larger cost for a project, estimated cost for 9 a project, might mean that the benefit-cost ratio is 10 unacceptable to government? 11 A. Absolutely. But with the inclusion of optimism bias 12 within that economic appraisal, effectively tests a more 13 onerous condition to allow funders or government to 14 proceed, and obviously there could be wider analysis 15 undertaken by government or funders to consider even 16 more onerous conditions. 17 Q. I would like you to look at something about optimism 18 bias in the preliminary financial case, please. 19 The document reference for that is TRS00000054. We 20 see the front page of the Preliminary Financial Case 21 here. Could we look, please, at page 27 of the scan. 22 Could we enlarge paragraph 4.4.1, both parts of it. 23 We can see you open that by stating -- I should ask, 24 did you contribute to this document? 25 A. Yes. 47 1 Q. You did. 2 "The Preliminary Financial Case has taken as 3 a starting point the fact that the project under 4 consideration represents a Standard Civil Engineering 5 Project, due to the number of other UK projects that 6 have been developed, and as a consequence the maximum 7 Optimism Bias adjustment to capital costs and works 8 duration is 44% and 20% respectively." 9 I think that's what you have just outlined to me 10 just now about why you chose that? 11 A. Yes. 12 Q. If you look at the second paragraph here: 13 "The project has now been in development for some 14 three years since its original promotion by WEL. As 15 a consequence of the stage it has reached, the analysis, 16 cross-checking and benchmarking of cost estimates, 17 together with the amount of mitigation that has been 18 carried out across the range of risk areas identified 19 above, it is considered appropriate to use lower factors 20 of 31% for Capital Cost Optimism Bias and 21 14% Works Duration Optimism Bias. It should be 22 noted that this compares to a capital cost contingency 23 of 11% identified by Mott MacDonald for Line 24 One. The difference between this figure and the 25 Optimism Bias Capital Cost adjustment of 31% 48 1 adds a theoretical GBP44 million of capital cost to the 2 project costs at this stage. As the project scope and 3 specification progresses, the risk mitigation exercise 4 should assist in reducing this number." 5 Now, if we could actually also look at the foot of 6 the page for the conclusions. At the bottom: 7 "An incremental Optimism Bias factor of 20 per cent 8 over and above the defined contingency as specified by 9 tie's consultants has been applied using HM Treasury 10 methodology." 11 That figure of 21 per cent there, was that the 12 difference between the 31 per cent optimism bias and the 13 11 per cent contingency identified by Mott MacDonald? 14 A. Yes. 15 Q. What did that 11 per cent represent, the Mott MacDonald 16 11 per cent? 17 A. I recall that that was effectively an estimate of 18 contingency that was necessary above their base costs. 19 Q. Now, that -- what you're doing there is deriving your 20 optimism bias percentage by looking at what would be 21 produced by the figure of 44 per cent and mitigating it 22 to 31 per cent, and then subtracting the contingency 23 allowance already made? 24 A. Yes. 25 Q. Why is the contingency allowance already made subtracted 49 1 from optimism bias? 2 A. When effectively you look at it this way, if you look at 3 Treasury guidance, it doesn't scale the costs of the 4 project and its contingency. It scales the base cost. 5 And therefore -- 6 Q. When you say scale, you mean the optimism bias is 7 applied to the base cost? 8 A. Yes. So it doesn't apply a correction to its 9 contingency. Because that effectively would be double 10 counting because you've got risks that are embedded in 11 optimism bias that may have a greater value or the value 12 that was estimated and then a broader set of risks, and 13 at these early stages, where you have less of an 14 understanding of all the risks that could be presented 15 to you in the project, you incorporate a larger headroom 16 within the optimism bias. 17 Q. But I can understand deciding that you're going to apply 18 the optimism bias figure only to the base figure and not 19 the contingency. But what you're -- here is not 20 deciding which figure you apply optimism bias to. 21 You're actually reducing the optimism bias by the whole 22 of the contingency before you apply it to the base, and 23 that's quite different, isn't it? 24 A. It's effectively ensuring that you're not double 25 counting. 50 1 Q. It's assuming that optimism bias and contingency overlap 2 100 per cent, isn't it, so that one can be deducted from 3 the other? 4 A. Correct. 5 Q. Is that borne out by guidance at the time? 6 A. Well, guidance at the time is -- is not crystal clear in 7 relation to that, all that aspects. But my 8 interpretation of it is that effectively you do not want 9 to double count and include risk allowances for the risk 10 twice. 11 Q. If you go back, please, to your statement -- 12 CHAIR OF THE INQUIRY: Could optimism bias have some 13 significance in the assessment of contingency? In other 14 words, if it's a psychological thing that you're being 15 optimistic, so that in assessing contingency, could you 16 be influenced by that psychological thing? 17 A. Yes, my Lord. That is one of the issues that subsequent 18 guidance has highlighted, that there could be a tendency 19 to be optimistic in your contingency allowance in the 20 knowledge that optimism bias is going to be 21 incorporated, or as a whole just be optimistic in your 22 contingency requirements, ie underestimating your true 23 risk exposure. 24 CHAIR OF THE INQUIRY: Are you saying that this is now 25 recognised in subsequent advice? Is that advice from 51 1 the Treasury subsequent to 2002? 2 A. This is from papers that have been produced by -- 3 I can't recall the name of the professor. 4 Bent Flyvbjerg, I think, who has effectively done 5 considerable more studies in the Dutch market with 6 regard to large scale infrastructure projects and 7 highlighted some of the broader behaviours that might be 8 inherent in projects. 9 CHAIR OF THE INQUIRY: Yes. He's the expert in this matter, 10 isn't he? He was the instigator of the Treasury 11 guidance. 12 A. He's one of the leading lights in this field. 13 CHAIR OF THE INQUIRY: Yes. 14 A. Without a doubt. 15 MR LAKE: I just want to look now at the assessment of 16 reductions to optimism bias, please. If we could look 17 at page 15 of your statement. If we highlight 18 paragraph 22, although it's -- I don't propose to read 19 through it here, but what you do is you note here what 20 was in the Treasury's 2002 guidance by way of the 21 ability to reduce the upper bound optimism bias as the 22 project proceeds. Is that fair? 23 A. Correct. In effect, the guidance, Treasury's guidance, 24 shows that as you progress to mitigate the risk 25 effectively identifying a mitigation factor, 52 1 a percentage complete of progressing that mitigation, 2 you can attach that to the various areas of optimism 3 bias and calculate a reduction. And the Treasury 4 guidelines have actually example calculations 5 incorporated within them that we followed and used as 6 a basis. 7 Q. Looking at the mechanism that you adopted to do that, if 8 I can jump forward to page 26 of your statement. If we 9 could highlight paragraph 32 at the foot of the page. 10 If you go down a bit. 11 You refer to a spreadsheet you prepared which is 12 available to all the parties. But if I can just read 13 after -- in the last four lines, you say: 14 "Reductions in optimism bias were made as progress 15 in mitigating risk reported by risk owners using their 16 individual judgement in relation to individual mitigation 17 factors for each risk." 18 Really, do I understand what you're doing here is 19 you're speaking to the people who had that risk and 20 getting them to assess the extent to which they've been 21 able to address it? 22 A. Correct. 23 Q. And adjusting the optimism bias factors accordingly? 24 A. Correct. 25 Q. Now, resorting here to the guidance within this, 53 1 employer, an Arup report in relation to this. Could we 2 look at production CEC01799560. 3 We can see that this is the front page of a 4 document, which was a review by Arup Scotland of the 5 Business Case that had been put forward. I'm just 6 wondering if there's a date on it. 7 In fact, if I ask you to go to page 30, we can find 8 a date on it there. If we look at the bottom right-hand 9 corner of the screen, I think we see the date of the 10 document. This date is 26 October 2004. 11 Are you familiar with this document, the review that 12 was carried out by Arup? 13 A. Yes, I am. 14 Q. Could we enlarge paragraphs 8.7 and 8.8, please. It 15 starts in paragraph 8.7 by noting what the uplift would 16 be. The conclusion of the first paragraph, having 17 considered the various percentile uplifts, and it's 18 looking at 57 per cent, which I think is the higher, 19 that's for the non-standard civil engineering; is that 20 correct? 21 A. No. They are making reference to DfT's guidelines. So 22 DfT did a specific study into a range of projects, and 23 then offered up different OB allowances for certain 24 projects. 25 Arup here are highlighting that if they were to 54 1 apply DfT's guidelines and -- which require -- and use 2 an 80th percentile for a rail project, bearing in mind 3 that this is a light rail project on a road, they would 4 take 57 per cent. 5 So it's not the Treasury guidelines and it's 6 effectively cross-referring to a heavy rail project, 7 rather than a light rail project. 8 CHAIR OF THE INQUIRY: We've heard evidence to the effect 9 that light rail projects can be more complex and 10 difficult to build than heavy rail projects which go 11 across green field sites or open countryside. Do you 12 have anything to say about that? 13 A. Actually, my Lord, the guidance shows that heavy rail 14 projects are more onerous to construct in terms of 15 having a larger outturn than originally planned. This 16 is driven by a number of reasons that are all captured 17 within optimism bias for that specific sector. Light 18 rail projects are considered to be less complex than 19 heavy. 20 CHAIR OF THE INQUIRY: Thank you. 21 MR LAKE: On the basis of the DfT guidance, they note at the 22 end there that on that basis, it might be considered 23 that the optimism bias uplift allowance may be 24 underestimated. 25 A. Yes. Effectively it draws a conclusion by pointing to 55 1 a different sector's delivery record and taking a more 2 onerous position. It highlights that if you take the 3 80th percentile for OB, that you could have the need for 4 a greater allowance. In effect, this is just saying 5 that, yes, circumstances could be very different, and if 6 you compared it to this other reference type of 7 projects, you could need more money. 8 However, when further on, they go on to highlight 9 the actual cost overruns in light rail itself, where in 10 the UK the experience has been that that's been up to 11 25 per cent. 12 So the guidance, bear in mind, always highlights 13 that your best starting position is to take a reference 14 class series of projects that are suitable to your 15 project, rather than taking broader, more onerous 16 approaches. 17 This was an issue that I considered right at the 18 start when appraising optimism bias, whether or not we 19 should be arguing that through detailed study of these 20 specific light rail projects and understanding that 21 25 per cent increase, we should be using that as more of 22 a guiding starting value than something coming from 23 Treasury. 24 Q. That was a 25 per cent increase, as we've said before, 25 over the final estimated cost? 56 1 A. Correct, correct. 2 Q. Not of the initial estimated cost? 3 A. Correct. Correct. 4 Q. Dealing with the question of when it's estimated, that 5 paragraph 8.7 notes at the end that the DfT guidance 6 recommends that its OB uplifts be applied at the time of 7 decision to build which is at Business Case submission. 8 It rather presupposes that's going to be quite early in 9 the overall project process. 10 A. Yes. This -- the preliminary financial case which Arup 11 are appraising here was a document that was prepared to 12 support the parliamentary process, effectively prior to 13 an outline Business Case production, where we would be 14 more developed in our design and understanding. 15 Q. Here obviously the intention was that the Business Case 16 submission and the decision was taken at a very late 17 stage, some contracts were already awarded. It was just 18 on the eve of the award of the Infraco contract. 19 A. The outline Business Case was prepared to allow us to 20 proceed with the earlier procurement such as the SDS, 21 Tramco, MUDFA, in knowledge that as we progressed 22 through the procurement of Infraco, there would be an 23 updated version of that, a Final Business Case prepared, 24 that would hopefully present the progress of those 25 advance works and commit to an actual price that had 57 1 been secured through the Infraco procurement. 2 Q. That Final Business Case was going to be the actual 3 decision to build. It was actually going to be very 4 late on in the process? 5 A. Correct. 6 Q. Do you consider that had an impact for how -- an impact 7 on the way in which optimism bias would be used in this 8 project? 9 A. I think in reality it was -- that period was of benefit 10 to us because effectively it allowed us to manage the 11 risks and deal with them as we progressed. We were 12 effectively unpacking a procurement strategy that was 13 de-risking the approach to delivery that these other 14 projects in the UK that suffered cost increases hadn't 15 taken the opportunity to deal with. 16 So detaching utilities, detaching the design and 17 dealing with progressing that -- an earlier stage, 18 bringing on board the operator and Tramco so that that 19 could be assembled and used to inform the Infraco was 20 all seen as an inherent better way of delivering than 21 other projects. 22 Q. If you could look at paragraph 8.8. If I read from the 23 fifth line, it says: 24 "However, DfT guidance also allows for downward 25 adjustments of optimism bias uplifts when warranted by 58 1 firm empirical evidence." 2 Now, you were adjusting down on the basis of the 3 impression of people as to how risks would be mitigated 4 in future? 5 A. Correct. 6 Q. Rather than on the basis of firm empirical evidence. 7 Was that not a departure from the forms of recognised 8 guidance for optimism bias? 9 A. I would highlight again that this is the DfT guidance. 10 Treasury's guidance doesn't have that type of limitation 11 on adjustment. It indicates that as you progress your 12 project, you can apply mitigation factors. The approach 13 that I took was to discuss that with each risk owner, to 14 understand their progress in mitigating that risk. 15 I saw the benefit of doing it that way was that you 16 weren't getting one individual guesstimating effectively 17 three or four parameters to calculate optimism bias, but 18 rather by taking a large group view across every risk in 19 the risk register, you are not going to receive a bias 20 to that. 21 So actually I felt as though our approach was 22 removing the risk of bias. It was dealing with that 23 issue that I think you're referring to that with the 24 individual quoting the mitigation factor, that they have 25 their own optimism bias in the value that they're 59 1 selecting. 2 MR LAKE: Thank you very much, Mr Bourke. Those are my 3 questions for Mr Bourke, my Lord. 4 CHAIR OF THE INQUIRY: I don't think anyone else has 5 suggested questions. 6 Thank you very much, Mr Bourke. That's all for the 7 present. You're still under citation, and technically 8 you could be recalled if something arose. Hopefully 9 that won't be necessary, but thank you very much. 10 A. Thank you, my Lord. 11 MR LAKE: My Lord, I was concerned the other witness wasn't 12 going to make it, but he is here. But this would be an 13 appropriate time to take a break. 14 CHAIR OF THE INQUIRY: We will resume again at 11.20. 15 (11.09 am) 60 1 INDEX 2 PAGE 3 MR MARK BOURKE (sworn) ...............................1 4 5 Examination by MR LAKE ........................1 6 7 MR JOHN CASSERLY (sworn) ............................60 8 9 Examination by MR LAKE .......................61 10 11 MR JAMES MCEWAN (sworn) .............................99 12 13 Examination by MR LAKE .......................99 14 15 Examination by MR DUNLOP QC .................174 16 17 18 19 20 21 22 23 24 25 193