1 Friday, 25 November 2022 2 (10.00 am) 3 MS KENNEDY: Good morning, Chair. 4 SIR WYN WILLIAMS: Good morning. 5 MS KENNEDY: Our first witness today is Mr Vincent Gaskell. 6 SIR WYN WILLIAMS: Thank you. 7 VINCENT GASKELL (sworn) 8 Questioned by MS KENNEDY 9 MS KENNEDY: Would you confirm your full name, please. 10 A. Vincent Gaskell. 11 Q. In front of you, Mr Gaskell, should be a copy of your 12 witness statement. Do you have that there? 13 A. Yes, I do. 14 Q. Have you read through this statement recently? 15 A. Yes. 16 Q. If you turn to the last page, is that your signature 17 there? 18 A. Yes, it is. 19 Q. Is it true to the best of your knowledge and belief? 20 A. Yes, it is. 21 Q. Your statement is now in evidence. Everything that 22 I ask you will be supplementary to that. Can I start by 23 thanking you for coming here and giving evidence to this 24 Inquiry. 25 I'm going to start by asking you a few questions 1 1 about your background. It's right, isn't it, that you 2 retired from the Civil Service in 2010? 3 A. That's correct. 4 Q. Prior to that, you held a range of roles within the 5 Civil Service? 6 A. That's correct. 7 Q. Those range from junior to senior management across 8 government; is that right? 9 A. Yes. 10 Q. And you held a role in the Benefits Agency? 11 A. Yes. Specifically in relation to this particular 12 Inquiry, yes. 13 Q. Yes, and that involved programme management for the 14 Agency's Customer Accounting and Payment System which is 15 referred to as CAPS? 16 A. Yeah. 17 Q. After that, you went on to work for the Home Office 18 before retiring; is that correct? 19 A. Yeah. 20 Q. Prior to 1997, you didn't have anything to do with the 21 Horizon project? 22 A. No, I didn't, or CAPS. 23 Q. But in 1997 you were asked to be programme manager for 24 the Benefits Agency CAPS? 25 A. Yes, that's right. 2 1 Q. Can you explain what that role involved? 2 A. Yes. I mean, briefly CAPS was a programme that was 3 initiated to bring together or initiate a common set of 4 personal details across all the Department's benefit 5 systems that had all been developed independently and 6 separately, and those systems also had their own unique 7 payment, ways of paying benefits at the end of it, in 8 terms of order books and giro cheques, but they had 9 their own modules for doing that. CAPS was about 10 bringing that together, producing a common module at the 11 back end, a common module at the front end and 12 implementing that across the benefit systems. 13 My role was quite specific to start with, in that my 14 job was to make sure all of that happened and to time, 15 to complete the replanning process that had been 16 initiated between POCL, ICL Pathway and ourselves, and 17 to just make that end of things happen. 18 Q. Prior to taking that role, did you have any technical 19 expertise with computer systems? 20 A. No, I had no -- I was not a technical person. What 21 I had was quite a bit of experience in technology 22 enabled change management, and I think it's for that 23 reason I was asked to take on this role, that involved 24 understanding what the technical capabilities were that 25 IT could help with, what it meant for the business in 3 1 terms of its impact on staff and on customers, and also 2 quite a bit of experience in managing multidisciplinary 3 teams of technologists, external consultants and 4 internal operational people. 5 Q. In early 1999, I think you say February 1999, you went 6 on to become programme director for the Benefits Agency 7 and the BA/POCL programme. So was that a promotion? 8 A. Well, no, it wasn't, except that I'd been gradually 9 brought -- the programme director, George McCorkell, had 10 been broadening my role in the run-up to the end of 1998 11 and I think it was a natural progression to just take on 12 the programme director's role, which I did. 13 Q. Can you explain what that role involved? 14 A. Yes, at that time, things had -- there was a lot of 15 discussion with the Department and with ministers about 16 future progress on the overarching programme, and my 17 role then was to be involved in those discussions with 18 the Department with ministers to try and find a way 19 through to make the programme work and, at the same 20 time, still keeping an oversight of the CAPS programme 21 to keep that on schedule. 22 I also then took over George's responsibilities for 23 some of the commercial aspects of the overarching 24 programme. 25 Q. So it's fair to say that between 1997 and 1999 you were 4 1 on a number of boards and committees involving the 2 Horizon project; would that be fair? 3 A. That's right, there was the CAPS programme board, which 4 I chaired, but then there was the CAPS and card 5 programme board, which George McCorkell chaired, that 6 involved ICL Pathway and also the Post Office, so that 7 took an overarching role -- view of the programme as 8 a whole. 9 The other board that I was on really was focusing on 10 CAPS itself. 11 Q. I think you also mention in your statement something 12 called the release authorisation board? 13 A. Yeah. 14 Q. Perhaps we can pull up an example minute of that, at 15 POL00028496. 16 Can you explain a bit about what the release 17 authorisation board involved? 18 A. Yes, I mean, typically, if there was to be a software 19 release or a major change, there'd be a -- something 20 called the release authorisation board, which was really 21 meant to judge the state of readiness of each of the 22 organisations or contributors to that particular stage 23 in the programme to gauge their readiness to -- for 24 implementation of anything that was under discussion. 25 Q. Thank you, that can come down now. 5 1 Turning back to the time you became programme 2 manager, I think when you start you say that the 3 interfaces and data links between CAPS and ICL were in 4 place and working; was that right? 5 A. Yes, that's correct. 6 Q. Can you explain what your feelings were towards the 7 project, or your first impressions, when you first 8 started in that role? 9 A. Right. I mean, at that time there'd been, as I'd been 10 given to understand, quite a bruising experience between 11 the three parties and my job then really, as I saw it, 12 was to keep the CAPS team completely focused on what 13 they were doing, to build strong relationships with the 14 programme design -- personal relationships with the 15 programme design authority, with my opposite number in 16 the Post Office, and also establish good relationships 17 with ICL Pathway. 18 Q. As part of your role, you had sight of the Horizon 19 business continuity reports; is that right? 20 A. That's correct. 21 Q. If we could pull up one of those at POL00028601, please, 22 and this interim report related to the period between 23 5 and 12 November 1997. 24 I think you say in your statement that these were 25 reviewed by the Programme Delivery Authority? 6 1 A. That's right. 2 Q. These interim reports, were they produced on a weekly 3 basis? 4 A. From memory, certainly initially, I remember seeing some 5 of these in the first few weeks of taking up the role. 6 As the programme developed, I think I saw few of these 7 and I can't remember whether these continued to be in 8 operation during the whole of 1998. 9 Q. But this particular -- if we look at this one -- 10 A. Yeah. 11 Q. -- between the 5th and 12th, this would have been 12 shortly after you assumed your role as programme 13 manager? 14 A. I think this is probably the first report I received. 15 Q. Really? Okay. 16 If we could turn to page 2, please, and scroll down, 17 we can see there that there was a grading system; is 18 that right? 19 A. Yeah. 20 Q. If we look over the page -- or can we scroll further 21 down that page, please. 22 If we look at 3.2, we can see under that "ABED" it 23 says: 24 "There has been an unusually high occurrence of 25 incidents at the ABED interface." 7 1 Do you remember what that stood for? 2 A. No, I don't, I'm afraid. 3 Q. But it goes on to say: 4 "This has been formally registered as a 'Problem' 5 and a solution is being sought from ICL Pathway. The 6 incidents have caused time delays and additional work in 7 the POCL Transaction Processing arena. This can be 8 contained in the current low volume environment but 9 gives cause for future concern." 10 So from the first report that you saw, issues were 11 being flagged? 12 A. That's correct, and, indeed, in each of the subsequent 13 reports, that was the case too. 14 Q. Perhaps if we look at the next report, which is at 15 POL00028600, please, and this relates to the period 16 between 13 and 19 November, so the following week. If 17 we go to page 2 of that report, and scroll down again -- 18 sorry, page 3, please, and scroll down. Thank you. 19 We can see again this issue being picked up and it 20 states: 21 "An item of deep concern is that ICL Pathway have 22 reported two incidences of duplicate payments being 23 made, however, this has not been evident from any 24 reports received, and ABED are awaiting further 25 details." 8 1 Was this something that concerned you from the very 2 beginning? 3 A. I think, from my point of view, given my quite specific 4 role on CAPS, what I was looking for in reports like 5 this were the view to be taken by POCL themselves about 6 the seriousness of these issues, and also of the 7 programme design authority, where the Department and the 8 agency had some very strong technical expertise. 9 I would also look to my own team to say "Are these 10 issues that we should keep an eye on, are we going to 11 keep looking for a pattern, are these things that are 12 just in the normal course of live running", as it turns 13 out in this case. 14 Q. For completeness, if we turn to the next interim report, 15 which is at POL00028599, this relates to the period 16 ending 26 November 1997. If we turn to page 4, please, 17 and scroll down, again we can see under one of the 18 bullet points, under 3.4.2, that there is an issue there 19 in terms of transaction processing and an error against 20 the cash account. Do you see that there? 21 A. Yes, I do. 22 Q. I appreciate that you say that you would look to people 23 in your team but isn't it self-evident, on the basis of 24 this report, that that would be a problem going forward? 25 A. Yes, when you say a problem going forward, what we're 9 1 looking for at this stage, we have a limited number of 2 cases that are live, we have a small number of post 3 offices that are live, and what we're looking for is 4 a number of things at this early stage in the programme, 5 is the seriousness of an issue like this -- and these 6 were serious -- the frequency of them and how quickly 7 they were being resolved, in this case by ICL Pathway, 8 and whose responsibility was it to make sure that that 9 happened. They're the sort of things we're looking for 10 at this stage. 11 Q. Turning then to a Horizon service report, which is from 12 December 1997, so shortly after that, POL00028596, and 13 these reports, were they done on a monthly basis? 14 A. The service reports, initially when I took on this role, 15 I remember them being almost weekly. Then I think there 16 seemed to be an increasing gap between these reports 17 being produced, but I can't swear to that. And 18 I certainly don't recall seeing reports of this nature, 19 say, by the middle of 1998. 20 Q. If we could turn to page 5 of that report, and if we 21 could look at the section "Lost Transactions". 22 A. Yeah. 23 Q. It says: 24 "There has now been 46 lost transactions, 8 of which 25 resulted in duplicate payments being made with a total 10 1 value of £254.75. The responsibility and liability for 2 these payments lies with ICL Pathway. 3 "ICL Pathway's preferred option for resolving this 4 problem, is to enforce the commitment of transactions at 5 the point of terminal timeout. The disadvantage of this 6 approach is that there is a chance that a transaction is 7 forcibly committed although the clerk intended to void 8 the transaction. If this were to happen then 9 a repudiation would inevitably be made by the 10 beneficiary who would be initially denied payment 11 because the system had falsely registered an encashment. 12 Historical data suggests void transactions are minimal." 13 So the same issue or similar issue is being picked 14 up again. Can you explain, in the grand scheme of 15 things you were dealing with, where did this land in the 16 pecking order of things that were on your mind or 17 concerning at the time? 18 A. I think from the -- and speaking personally -- well, 19 I'll do both for the Agency and for myself. When I saw 20 issues like this, what was of immediate concern to me 21 when I saw that there were duplicate payments and these 22 were potentially relating to benefit payments, is the 23 impact on the customer: would they be asked to refund 24 the amounts of money that had been overpaid by way of 25 a duplicate payment? So there is an issue there about 11 1 customer service and the impact on individuals. 2 There is an impact then on the integrity of the 3 system and -- in terms of what we're looking at. So, in 4 that case, what I'd be looking to is our finance 5 colleagues, both in the Agency and the Department, to 6 say "Is this something at this point in time we should 7 worry about, is this something that we're content that 8 ICL Pathway are getting on top of to resolve, are we -- 9 again, are we seeing a pattern?" 10 In this case, in terms of duplicate payments there 11 was a period in which we saw some duplicate payments, 12 but, if I recall from mid-1998 onwards, that was less 13 evident, if -- I don't recall any after that. 14 Q. You mention or we mentioned previously the Programme 15 Delivery Authority and I believe you say that you 16 attended some meetings from the beginning as 17 George McCorkell's deputy. 18 If we could turn one of those up at POL00028310, we 19 can see your name there. 20 Would you often attend these meetings as a deputy 21 for George McCorkell? 22 A. No. No, George usually took this role on himself. 23 Q. How did you find those meetings? 24 A. I think they were quite challenging, is the best way to 25 describe it. I think this is where the -- some of the 12 1 more serious issues on the programme, both in terms of 2 timetabling and any issues that were coming up through 3 programme delivery, would actually be challenged at this 4 senior level. So I would say these were -- I think the 5 best way I can describe it is challenging: professional 6 but challenging. 7 Q. Various witnesses have said that they found this process 8 frustrating or difficult. Did you sense that when you 9 attended these meetings? 10 A. No, I didn't. I actually found this is where things got 11 a real focus, in terms of whether things were actually 12 being done to schedule and were there any significant 13 issues arising. When you're in a forum like that and 14 those issues are coming to the fore, where people are 15 actually feeling challenged, I can see how others would 16 find that uncomfortable. 17 Q. Turning then to issues regarding testing, I think you 18 say that there were disagreements about testing 19 requirements between BA and POCL; is that right? 20 A. When we got to middle to late 1998, that was the case. 21 Not in the early stages but in middle to late 1998, we 22 had a different approach and a different thought 23 process, is the best way I can describe it, as to the 24 extent of testing needed and how testing should be 25 conducted. 13 1 Q. If we can pull up a board minute from that time, 2 POL00028395. So this is a "CAPS and Card Programme 3 Board Action Notes" minute -- 4 A. Yeah. 5 Q. -- from 12 June. 6 If we turn to page 2, we can see under "Detailed 7 Testing Requirements": 8 "There are still a number of uninvolved differences 9 between BA and Pathway over the Detailed Testing 10 Requirements. Dave Miller explained that there was 11 a need for a senior level meeting to determine exactly 12 what the testing requirements are going to be. Peter 13 Crahan said that the issues should be tabled at the 14 Formal Management Review Meeting on 18/6/98. The Chair 15 sought and received assurance that if this was not the 16 case, escalation routes were in place and that there was 17 the facility to resolve this issue." 18 So is this around the time that these issues start 19 appearing and are starting to bite? 20 A. Yes, I think there was a quite -- I'm going to use the 21 term "philosophy" when it comes to approach to testing. 22 We were certainly, having been involved in considerable 23 number of system developments within the Department, we 24 had a particular thought process about what testing was 25 needed for a large scale development of this kind, and 14 1 that view was not shared, it's fair to say, by both POCL 2 and by ICL Pathway. 3 So, therefore, trying to agree -- and I don't mean 4 that in any critical way of anybody, we just had 5 a different approach. And we felt that the testing 6 should be more comprehensive and thorough, in our terms, 7 than looked to be scheduled. So that's where we started 8 to begin to part company on testing. 9 Q. In your view, were those differences fundamental? 10 A. I think they were, and I think that -- that came to 11 a head in early 1999. But yes, they were quite 12 fundamental. 13 Q. If we could turn to page 6 of that same document, 14 please, this is "Input into CAPS Programme Board Report 15 for Meeting on 10 July 1998", and if we could look, 16 please, at "Live Operations", we can see that it says: 17 "The current Release continues to provide a Benefit 18 Payment Service and Order Book Control System for Child 19 Benefit Payments in 204 outlets (119 in the South West & 20 South Wales Region and 85 in the North East Region). 21 From an operational point of view, there are no major 22 threats to business continuity. 23 "During May, there were approximately 60,000 24 encashments (including foreign encashments) with a total 25 value in excess of £2 [million]. 15 1 "However there are concerns that current levels of 2 lost transactions (at 4 per 10,000 encashments), 3 although presently manageable, may not be acceptable to 4 POCL when large volumes of encashments are returned by 5 newly automating post offices. Reducing the level of 6 lost transactions therefore remains a prime focus for 7 POCL Service Management. The Priority Level 1 assigned 8 to these problems is focusing the Contractor towards 9 activity addressing POCL concerns and providing remedial 10 action." 11 So, at this stage, was this really POCL's concern 12 regarding lost transactions, rather than BA? 13 A. Yes, it was, and that was, as you -- in some of the 14 other papers that I was sent for this Inquiry, that was 15 increasingly the case during 1998, where POCL were 16 increasingly concerned about the number of lost 17 transactions, and that's -- that featured at most of the 18 reports that I saw. 19 Q. If we turn to page 7 of that same document, please, and 20 we scroll down, we can see the "Testing" there under 21 "ICL Pathway Release 2". Under the second paragraph it 22 says there: 23 "... some concerns and issues that need to be 24 resolved if End to End and Model Office testing are to 25 start on 3 and 10 August respectively ..." 16 1 So both the testing issue and the lost transaction 2 issue are coming to a head at this stage? 3 A. Absolutely. 4 Q. If we could look at the Horizon service report from, 5 again, a similar time, which is at POL00028589, and if 6 we turn to page 6, we can see, in relation to the 7 Benefit Payment System, that "Incomplete Transactions" 8 is given a whole page for discussion here, isn't it? 9 A. Yes, it is. 10 Q. Is your memory again, at this stage, this is a POCL 11 issue rather than a BA issue? 12 A. I think there was a lack of certainty about what the 13 lost transactions were. As I recall, and my memory may 14 be faulty here, but if -- as I recall, during the course 15 of 1998 there was much discussion about what these lost 16 transactions were. I think ICL Pathway had a view that 17 some of these, if not a lot of them, were down to user 18 error. I think the Post Office contested that, and, 19 from our point of view, we just wanted resolution, 20 because we were worried about what would happen when we 21 began to scale up the system during 1999. So this 22 became quite an issue during 1998. 23 Q. If we turn to page 8 of that document, and can we scroll 24 down and perhaps flip the graphs so we can read them. 25 This is the breakdown of the incomplete transactions 17 1 that were reviewable at that time; is that right? 2 A. Yeah. 3 Q. If we turn over the page to page 9, we can see again 4 under "Incomplete Transactions": 5 "The transaction Processing & ICL Pathway interface 6 remains on 'Amber' status, with incomplete transactions 7 having a large impact on this. 8 "Other factors which are of concern are: 9 "unmatched exceptions 10 "cash account errors. 11 "Last months report included a statement from TP 12 detailing the impact of incomplete transactions within 13 their domain, in addition, a report has been furnished 14 showing the projected effect in monetary terms to the TO 15 environment." 16 A. I think it's at this time that it prompted me to ask 17 ICL Pathway for a detailed breakdown of what they knew 18 about lost transactions at that point. I think it was 19 at this point, rather than later. And ICL Pathway 20 provided such a summary, which included what had been 21 done to resolve earlier issues and what outstanding 22 fixes and subsequent actions were needed to resolve any 23 remaining issues as they saw them at that time. 24 Q. If we turn to page 23 of this same report, under 25 "Concerns", we can see: 18 1 "Lost transactions form 70 per cent of incidents 2 received this month. Concerns therefore remain and 3 service management is specifically tasked towards 4 resolving the underlying causes. An action group has 5 been specifically targeted at this for resolution." 6 At this stage, were you still of the belief that BA 7 would proceed with the project and be party to it? 8 A. Yes. 9 Q. In spite of all of these issues, you thought that -- 10 A. Absolutely, and I can say that the CAPS team, in 11 particular, were entirely focused on making that happen. 12 That was clearly in our mind that that's what we were 13 going to do. 14 Q. If we move forward, then, to November 1998, if we could 15 pull up POL00028433, and if we could zoom in a bit, 16 please, and down. This is a letter that you wrote to 17 David Miller at POCL, and you tell him that there are 18 three issues that are outstanding. You can see the 19 first at the bottom of that page, which is about the 20 provision of test condition analysis for end-to-end 21 testing. 22 Then if we go over the page, the "analysis of PinICL 23 clearance and prioritisation" and, thirdly, 24 "formalisation of the weekly progress meetings and 25 pre-run checkpoint process for the start of testing", 19 1 which you describe as "absolutely essential before we 2 enter the final runs of testing". 3 At this stage, what is the relationship like between 4 you on the one hand and David Miller on the other, or 5 POCL and BA? 6 A. I would have said we -- David -- I had a very good 7 working relationship with David Miller. It was 8 professional, again challenging, when we needed to be 9 challenging of each other, but thoroughly professional 10 and I thought it was a good relationship to the point 11 where we could phone each other up if there was an issue 12 that was brewing that we wanted to forewarn each other 13 about, might come up in a meeting. It would be that 14 type of relationship but testing, challenging. 15 Q. In the last paragraph of this letter you say: 16 "I also understand that during current phases of 17 testing significant problems are being experienced with 18 the TIP interface and that investigations are underway 19 to determine their severity. Could you confirm that my 20 understanding is correct and if so what proposals are 21 being considered to address the problems and how will 22 they compact on the current phases of testing and on 23 overall progress towards the NR2 release. It would be 24 helpful to have a response to this particular set of 25 issues before 13 November." 20 1 Would you say that reflects what you've described, 2 the difficulty with testing? 3 A. Yeah, it was, and we were getting early indications that 4 things weren't going well. We'd had early indications 5 that there were timetabling issues, we'd had early 6 indications that the testing was not going well, which 7 is actually just fuelling our concerns about the state 8 of progress at that point. 9 Q. If we could turn to page 3 of that document, thank you, 10 and if we could scroll down, this is Mr Miller's 11 response to you on 10 November 1998, and he takes each 12 of these three issues in turn. If we look at, over the 13 page -- if we look at the final bit in relation to TIP, 14 it says: 15 "The problems you refer to with regard to the TIP 16 interface are of concern. Simon Rilot has made that 17 issue his first priority, in order to resolve any 18 problems that may prevent POCL and BA having confidence 19 in the end to end accounting and reconciliation process. 20 He has organised workshops later this week in 21 Chesterfield and you should be aware that the output of 22 those workshops may change the shape of the final phases 23 of testing. You can be assured that Carol will be 24 directly involved in these discussions." 25 So, at this stage, there is still quite a lot of 21 1 back and forth, trying to resolve these issues; is that 2 right? 3 A. Yeah, that's correct. 4 Q. If we could then turn up the Horizon programme's issues 5 register from December 1998, which is at POL00028377, 6 and if we could zoom in on 10003, it says: 7 "[The] Inability to achieve satisfactory cash 8 account balancing." 9 If we also look at 10024, we can also see that 10 there's a reference to the EPOS element. Thank you. 11 So this risk register, was this something that you 12 generally had sight of? 13 A. I can't specifically recall. I think I probably would 14 have, but I can't specifically recall. 15 Q. If we look over at page 3, please, and if we could zoom 16 out a bit, thank you, the bottom email we can see 17 an email from Sue Muddiman(?), do you remember who that 18 is? 19 A. Yeah. 20 Q. It says: 21 "Below is a list of the questions that Dave Miller 22 may be asked by Vince Gaskell at the CAPS and Cards 23 Programme Board ... 24 "We need to provide the answers for Dave so he can 25 be prepared so I suggest that this is passed to the 22 1 attendees of the HMT meeting tomorrow for that very 2 purpose." 3 It goes on to list a number of issues, and it's 4 a number of things that I think, if we scroll back up, 5 we can see that you did ask for these things. Can you 6 explain a bit about your thinking at the time as to why 7 you asked for these things? 8 A. Yeah, I think the issue continued to rumble on about the 9 nature of the testing, the scope of it, definitions for 10 entry and exit criteria from testing to judge whether it 11 was a success, all of these building towards whether 12 things were acceptable or not, so what was the 13 acceptance specifications that we'd be looking for. And 14 the absence of those or the lack of clarity of those was 15 quite a worrying thing because how would you know that 16 the testing had been successful, unless you'd actually 17 agreed those upfront, knew what outputs or outcomes you 18 were looking for. Those things still at this seemingly, 19 to me, late stage were still outstanding, so that's why 20 I think these were being raised in my name because these 21 were quite an issue for us. 22 Q. If we look at number 2, we can see you raise the issue 23 of inability to achieve satisfactory cash account 24 balancing. So, at this stage, would you say that comes 25 within the scope of those things you were saying -- 23 1 A. Oh, yeah. 2 Q. -- at this late stage should have been resolved? 3 A. Yeah. 4 Q. If I could ask you to look at the second Project Mentors 5 report -- I think actually it's the third one. It's at 6 POL00031114. If we could scroll over -- well, actually, 7 staying on that page, we can see that this was sent to 8 George McCorkell, who would have been your boss at the 9 time? 10 A. Yeah. 11 Q. If we scroll over on to page 2, and down, did you 12 receive a copy of this report at the time? I appreciate 13 we only sent this to you last night and gave it to you 14 this morning, but -- 15 A. No, I didn't, and indeed, on the front page, it does say 16 that it's not to be shared beyond the addressees. So 17 no, I didn't receive a copy of that. I think before 18 I took on the role of programme director from George, 19 George gave me a verbal briefing and mentioned this but 20 I never saw the report. 21 Q. When he gave you that verbal briefing, did he tell you 22 the content or the gist of what the report had said? 23 A. The gist, yes. 24 Q. So you were aware of its critical findings? 25 A. Absolutely, which tended to reinforce my view about what 24 1 we needed to do on the testing front. 2 Q. So you would have become aware of this report in 3 February, when you took over -- 4 A. Yes, that's correct. 5 Q. Turning back, then, to January 1999, if we could pull up 6 POL00028410, and if we could turn to page 3, in the 7 middle of paragraph 8, there's discussion about 8 functionality, and then you're recorded as having said: 9 "Vince Gaskell said that before the CAPS Programme 10 agreed the entry criteria for End-to-End testing and 11 Single Benefit Model Office an assurance was required on 12 BES functionality, it was agreed that there was a need 13 for further discussion on this subject outside of the 14 meeting." 15 Then if we scroll further down onto page 6, 16 please -- sorry, staying on that page, paragraph 12, we 17 can also see that you've said: 18 "Vince Gaskell said that it was the CAPS Programme's 19 perception that Cycle 1 had not gone well because it had 20 generated 20 high/medium PinICLs which had not arisen in 21 the earlier stages of testing. Andrew Simpkins said 22 that some PinICLs had been expected and that they would 23 all be cleared before Cycle 2." 24 At this stage, are you starting to feel like things 25 are not going to improve or were you still hopeful that 25 1 they would? 2 A. No, I was -- we wanted this thing to improve. What we 3 were concerned about is that there were signs that 4 things weren't going well and that each of these cycles 5 of testing where we were expecting less PinICLs to be 6 generated, then that's not what we were seeing. So we 7 never got to a stage where we saw what we would regard 8 as a clean run of testing. 9 Q. Turning then forward to February 1999, so this is when 10 you've taken over as programme director, if we could 11 pull up POL00028408, please, this is a CAPS programme 12 board report and, if we scroll down, please, we see 13 here, again, the raising of the issue of incomplete 14 transactions: 15 "Historically Incomplete Transactions have been 16 reported a month in arrears due to the frequency of 17 ICL Pathway's reports. However, because of their 18 importance the current status of ITs will now be 19 reported based on the latest information produced by the 20 BSM Incomplete Transaction Workshop." 21 So again at this stage, this late stage of 22 February 1998, this is being reflected in the notes of 23 the board meetings? 24 A. And what we were looking for there, I mentioned we were 25 looking throughout the process about trends, and there's 26 1 a reference here to what had happened the previous 2 month, and the numbers in the previous months. So these 3 were things we were just keeping a careful eye on. 4 Q. If we turn over the page to page 2, then scroll down to 5 the bottom, we can see that testing is raised again as 6 an issue, and if we scroll on to the next page, can we 7 have two pages -- yes -- saying that this needs to be 8 resolved, essentially? 9 A. Yeah. 10 Q. Moving then forward to March 1999, there was a meeting 11 between BA and POCL teams, and if we turn it up on 12 HMT00000018, and we can see that you were in attendance 13 there. 14 If we scroll on to the second page, it mentions that 15 you're feeling positive about a new approach. Could you 16 explain whether that reflected your attitude at the 17 time? 18 A. I think that rather overegged it. I think what I was 19 desperately keen to do is that there had been quite 20 a lot of discussion already taking place about the 21 future of the programme, and I'm more action orientated, 22 and this was an opportunity to say how could we at least 23 try and get a grip on one aspect of this, which is what 24 they were referring to in this particular paragraph. 25 So this was an opportunity to do some work that 27 1 might help us progress to a point where we could make 2 a decision on a way forward. That's what I was keen to 3 do and I think that's what was reflected in my 4 enthusiasm to take up this approach. 5 Q. In April 1999, it was decided that BA CAPS would not be 6 undergoing the live trial, and this decision was 7 confirmed, if we turn up POL00028405. Can you explain 8 a bit about why that decision was taken? 9 A. Okay. Just first of all, to explain Val Curran actually 10 took over the programme manager's role from me, working 11 to me as the programme director. So this was where we'd 12 had increasing concerns -- as reflected in the CAPS and 13 card programme board minutes, that you've already 14 alluded to, that George McCorkell chaired in January -- 15 where issues to do with what we had seen from the 16 evidence from testing was not good. We'd not got to 17 a stage where we'd seen a clean run of testing, and yet 18 we were being asked by POCL and by ICL Pathway to 19 progress to the next stage, which actually quite worried 20 us, because we were then worried about what did that 21 mean for timetable -- the timetable overall. 22 I'll start with that point first of all, because the 23 timetable had not been adhered to, even since the 24 re-plan in 1997. So we were concerned then about what 25 that would mean for the timetable for work on our major 28 1 benefit systems that would need to be involved for the 2 multi-benefit stage during 1999. That was the first 3 point. 4 The second stage, and perhaps most crucially, we 5 didn't have confidence that sufficient testing had been 6 done to make sure that we weren't going to experience 7 further problems and, therefore, further delays during 8 the next stage of testing. It's a difficult one for us 9 to conceive of a situation in which you would progress 10 to the next stage of a live trial and possibly towards 11 acceptance because -- possibly towards acceptance and 12 therefore exchanging large amounts of money, public 13 sector -- public money into ICL Pathway, if we weren't 14 certain that it was the right decision to proceed and we 15 were far from certain that it was the right decision to 16 proceed, regardless of all the other considerations that 17 were taking place way above us in terms of ministerial 18 discussions. 19 This was at a programme level: was it sensible to 20 progress with the timetable that was being put on the 21 table, given what we had seen from the earlier stages of 22 testing and the lack of resolution of some of the 23 outstanding matters. 24 One specific is that, to move from one stage of 25 testing to another, even though we were assured that the 29 1 PinICLs were being resolved during the previous stage of 2 testing, that might have been the case but you get to 3 the next stage, you worry whether the resolution of 4 those PinICLs has actually undone something as you move 5 to the next stage of testing. 6 So that's why we wanted to get to a point where we 7 could see a cleaner run of testing, which we never got 8 to, and therefore we concluded it was more realistic to 9 pause and that's why we'd decided not to progress. 10 That decision was not taken by Val Curran, it was 11 taken in discussion with myself with the Agency chief 12 executive, with the Department's Permanent Secretary and 13 also with the Secretary of State. 14 Q. It's been suggested that one of the reasons why the BA 15 didn't undergo the live trial was because it knew that 16 the DSS was halfway out the door and didn't fully intend 17 to go through with the project. What would you say in 18 response to that? 19 A. I would say I don't accept that, certainly not at the 20 level I was working at, and certainly not with my 21 intention to try and make this programme work. 22 The efforts that we'd gone to, to work with POCL to 23 try and resolve the outstanding matters in terms of 24 testing, the efforts we'd gone to with ICL Pathway to 25 adjust our dates to try and make things happen, none of 30 1 that is -- all those efforts -- none of that is 2 consistent with that point of view, so I reject that 3 point of view. 4 Q. If we could turn up POL00028406, please. This is 5 a letter from David Miller to you on 8 April 1999 6 dealing with issues that you've raised in respect of the 7 project. But if we could turn over the page to page 2, 8 the "Stability of the Solution", and looking at the 9 first paragraph, he says: 10 "We cannot understand the assertion that the 'level 11 of risk of new faults arising is unknown'. Clearly 12 there is progressive evidence from each of the test 13 cycles and acceptance reviews. We cannot of course be 14 certain that no new faults will arise but that is the 15 nature of the process. The CAPS/Benefit Payment 16 System ... areas have performed consistently as testing 17 has progressed, with the later faults arising from 18 specific conditions and not from underlying design 19 flaws. We believe the BA itself recognises that the BPS 20 elements of the system are stable." 21 Was that your position? Would you have accepted 22 that those aspects of the system were stable? 23 A. Yes, I think I would. I think it was just the wider 24 picture that we were concerned about, and the scale of 25 issues facing POCL and ICL Pathway in particular, from 31 1 the test results, so -- and we were therefore about what 2 impact that might have on the quality of the system that 3 was being -- as we entered into live trial, and what 4 impact that might have on timetable subsequently to put 5 things right. 6 Q. What was your relationship with Mr Miller like at this 7 specific time, April 1999? 8 A. Still I would say quite professional. I think we both 9 were able to express disappointment that we couldn't 10 agree, but that's quite natural in a situation like 11 this. I would still say it was professional and good 12 natured. 13 Q. On 11 May 1999, you proposed that Child Benefit 14 customers be removed from the Benefit Card Payment. If 15 we could turn up DWP00000007, yes, that's the one, and 16 if we could turn to page 6, please, and scroll down, 17 please. This is a memo that you wrote to Bruce McNiven 18 at the Post Office and Tony Oppenheim at ICL, and you 19 say: 20 "Further to discussions last week regarding the 21 removal of Child Benefit Customers from the [Benefit 22 Payment Card], I undertook to provide you with some 23 further details on the CAPS view of how this might be 24 performed. 25 "In order to successfully remove all customers from 32 1 Benefit Card Payment, ensure continuity of benefit 2 payment and avoid adverse publicity full co-operation 3 would be required from ICL Pathway, POCL and DSS. In 4 order to be fully prepared, should a decision be taken 5 to remove customers from card payment, we would 6 recommend that early meetings are held with all three 7 parties to: 8 "understand the overall framework for the removal of 9 customers; 10 "develop and agree the detailed processes which need 11 to be undertaken; 12 "develop and agree a testing and implementation 13 plan, which would be used to remove customers from 14 card." 15 So at this stage what's happening? What discussions 16 are happening in relation to the Child Benefit customers 17 being removed? 18 A. I think -- I'm trying to recall where this sat with the 19 discussions between ministers and -- of both departments 20 and with the Treasury. But, clearly, if cards were not 21 going to continue, then we would need to work out what 22 we needed to do about existing customers on Child 23 Benefit. So, again, this was a preparatory piece of 24 work to understand what we would need to do. 25 Q. If we turn back to page 1 of that document, and this 33 1 records the Benefit Payment Card being cancelled, how 2 did you feel about that at the time? 3 A. It was mixed feelings, to be honest, because I was 4 relieved, the programme team were relieved, we'd got 5 a decision, and we could move on to, towards using what 6 we -- the work we'd already gone to prepare for ACT. 7 But disappointment that things had come to had a head 8 like this. 9 Q. If we could turn up POL00028730, this is a letter that 10 you wrote to -- sorry -- to John Bennett, and if we read 11 the first paragraph, we see: 12 "We reject your allegations that we are in breach of 13 our obligations to Pathway under the Related Agreements. 14 We do not agree that all of the CCNs listed in the 15 schedule to your letter are waiting for a response from 16 us." 17 What is the relationship like at this stage? 18 A. With ICL Pathway? 19 Q. Yes. 20 A. I think that it was quite difficult by now. I think 21 they were understandably feeling pretty sore about the 22 decision that had been taken to cancel the Benefit 23 Payment Card and pretty sore about the -- I can't 24 remember at what point we had reached in trying to get 25 a settlement of commercial matters with ICL Pathway at 34 1 this point, but they were pretty sore about the way 2 things were looking. 3 Q. If we could turn to page 3 of that document, and scroll 4 down. Sorry, this is the letter that you wrote to 5 Bruce McNiven, which states: 6 "As I indicated to you during our [phone] 7 conversation yesterday, the reports I have received 8 following the Data Centre Migration differ to the 9 information you have now provided in your letter about 10 the scale of the problems that have arisen." 11 What was the scale of the problems at that time? 12 A. A number of -- I can't remember the specific number, but 13 it was a significant number of Child Benefit customers 14 had been affected by the data centre migration. I think 15 it was affecting the likelihood that they would not have 16 continuity of payment through the Benefit Payment Card, 17 so we needed to take remedial action to put that right. 18 Q. Around this time, KPMG prepared a report, if we turn up 19 HMT00000008, thank you. If we look at page 5, and if we 20 scroll into and down, please, we can see that you 21 contributed to this report. Can you explain what that 22 involved? 23 A. Yes, I mean, the -- by this stage, the Department and 24 other government departments were heavily involved in 25 working out what to do next, not just in terms of the 35 1 Benefit Payment Card but what it would mean for 2 continuity of payment by order book in the short-term 3 and movement towards ACT. So my role, therefore, on 4 behalf of the Benefits Agency, was to co-ordinate input 5 to that, some of which was more data specific in terms 6 of flows of -- and projected flows of order book 7 payments, and so on. So anything technical about CAPS 8 and what we would need to do was input to that too but 9 this was more about supporting the Department and the -- 10 and Government ministers on trying to find a way 11 forward, or to provide them with information that would 12 help with the options that were under consideration. 13 Q. Subsequently, you received various letters from 14 Mena Rego at the Post Office. If we could turn up 15 NFSP00000041, please, and if we could scroll down. 16 Sorry, if we could go to page 6 of that document, 17 and scroll down, please. 18 This is a letter that she wrote to you regarding 19 acceptance, and it says: 20 "Following our discussions on Acceptance on 21 Wednesday 23 June -- when I believed we were on the same 22 wavelength -- I was not a little surprised to hear from 23 my contract team that we had now received a draft 24 schedule 16b. This seeks (amongst other things) to 25 impose a full blown Acceptance process on POCL for the 36 1 OBCS service which effectively replicates and enhances 2 the previous arrangements between POCL and ICL. 3 "Given the stage we are at on acceptance with ICL, 4 and bearing in mind your own Team's heavy involvement 5 (and Leadership) on the definition and development of 6 the solution as well as the early stages of the 7 Acceptance process, I believed we had agreed that the 8 way forward was to carry on informing and consulting you 9 regarding the OBCS service whenever appropriate." 10 Can you explain what's going on here? 11 A. Yes, I think Mena and my memory of what we'd discussed 12 clearly diverged. We were still looking for acceptance, 13 understandable acceptance, for what was to happen with 14 the Order Book Control Service and how we were going to 15 be satisfied that was going to be working okay in the 16 short term. 17 Q. What did you think about the Horizon project going 18 forward and whether or not it would be viable for POCL? 19 Was that something that you thought about at the time, 20 after BA withdrew? 21 A. I think things that I'd discussed with Dave Miller and 22 others were that, on the back of the view that we took 23 about the state of readiness to move into live trial, my 24 concern for them was whether they were going to continue 25 to face ongoing issues of timetabling and also quality 37 1 of the software that they were going to be operating. 2 MS KENNEDY: Thank you. 3 Chair, I don't have any further questions for 4 Mr Gaskell. Do you have any questions at this stage? 5 SIR WYN WILLIAMS: No, I don't, thank you. 6 MS KENNEDY: Sir, I'm not sure whether any of the core 7 participants have questions. 8 Yes, Mr Stein does. 9 SIR WYN WILLIAMS: Right. Over to you, Mr Stein. 10 Questioned by MR STEIN 11 MR STEIN: Thank you, sir. 12 Mr Gaskell, my name is Sam Stein and I represent 13 a large number of ex-subpostmasters, mistresses and 14 managers. 15 I will take you, if I can, please, straight to 16 a document, which is FUJ00075730. We should be seeing 17 there, as you can see on this page, this is a document 18 saying: 19 "Presentation/Demonstration to [Right Honourable] 20 Frank Field MP, Minister for Welfare Reform [and then 21 the date] 16th July 1998." 22 You should have been provided with, I think, some of 23 these materials beforehand and hopefully you have had 24 an opportunity to read them? 25 A. I only saw them first thing this morning. 38 1 Q. In that case, I will take you slowly to the point I'm 2 about to make. 3 Can we go to page 18 within the Relativity 4 pagination. Thank you. If we look at this document, 5 which, you see at the top, it says "Eyes on The Future", 6 the bottom strapline says "Feet On The Ground". Under 7 "Fraud Savings", it says this: 8 "Positive authorisation of all payments 9 "Counterfeit/lost/stolen cards 10 "Cardholder verification 11 "Pick Up Notices ... 12 "Extended Verification Procedure ... 13 "Signature and card detail checks. 14 "Fraud Risk Management Service 15 "Monitoring, trend analysis 16 "Investigation support." 17 So we can see what's being discussed here, which is 18 the need for particular aspects of the system in 19 operation. Now, this appears to be in reference to the 20 time period when we're looking at the involvement of BA, 21 the Benefits Agency, yes? 22 A. It does look that way, because we're talking about, 23 really, cards and replacing paper-based methods of 24 paper. 25 Q. Exactly and that was the desire of the Benefits Agency, 39 1 which was to move to ACT, the card system; is that 2 correct? 3 A. Well, it's to move to the card system, in the first 4 instance, and ACT eventually. 5 Q. Yes. Now, my question in this aspect of things is that 6 we can see that this relates to the Benefits Agency, we 7 can see what's being said about fraud savings, we're 8 aware, or the Inquiry is aware, that there was 9 a considerable concern within the Benefits Agency about 10 millions being lost through fraud, yes? 11 A. Absolutely. 12 Q. Right. So let's look at the other side of things. What 13 information did you have at that time about the Post 14 Office's prosecution service or Post Office prosecutions 15 more generally? 16 A. I, personally? None. 17 Q. Were you even aware that the Post Office prosecuted its 18 own cases? 19 A. Only as a general member of the public. 20 Q. Were you aware that the Post Office had its own 21 investigation team? 22 A. No. 23 Q. Thinking back, can you help, then, with this: what 24 discussions were you involved in or knew about between 25 the Post Office, Pathway and Government? What 40 1 discussions were ongoing about the prosecutional needs 2 of the Post Office? 3 So this is the BA, the Benefits Agency; what was the 4 Post Office saying about their issues with 5 investigations and prosecutions? 6 A. I'm not aware there were any and, just from memory, 7 thinking back to the CAPS and card programme board, 8 which is where the overarching programme was reviewed by 9 the Agency, that was never an issue that came up, that 10 I can recall. 11 Q. Now, the same question, just departing slightly from 12 prosecutions, which I've said so far. The Post Office 13 also deals with its own civil cases in relation to 14 monies that they decide that they should, at that time, 15 seek to recover through the civil courts. Again, what 16 discussions were there from the Post Office as to its 17 needs regarding the use of material from the Horizon 18 system for civil actions? 19 A. None that I'm aware of. And I can only -- just in order 20 to try and be helpful, when the issues cropped up about 21 incomplete or lost or duplicate transactions, issues 22 like that were never mentioned. 23 Q. If we can cap this off, then, did you ask any questions 24 about it? They weren't mentioned but did you say "Well, 25 what's going on with your side of all of this?" 41 1 A. Only insofar as what was the impact of the lost, missing 2 and incomplete transactions, where the suggestions were 3 being made that this could be down to user error by 4 subpostmasters. That was where the focus was. 5 Never was it suggested or even mentioned, for 6 instance, that there might be something inappropriate 7 taking place. This was more about whether the system 8 was performing or whether the staff who were operating 9 the system in sub post offices were doing it in the 10 right way. 11 Q. All right. So if summarise this, then, essentially, you 12 don't recall there being any discussions from the Post 13 Office about the needs of their own systems for 14 investigations for civil actions or prosecutions? 15 A. No, at no stage. 16 Q. I'll take you to a separate document, please, which is 17 POL00028530. Now, we will see this one at -- if we can 18 just go down to, I think it's the second page, yes, 19 there we go, right. 20 Does this come under the same heading of a document 21 you looked at this morning or had this morning? 22 A. You'll have to show me more, I think. 23 Q. You can see what this is, "BA/POCL Automation Project, 24 Interdepartmental Working Group Report To Ministers", 25 okay? So the date of this is 13 November 1998, so if 42 1 that helps orientate ourselves. 2 Can we go to the top of page 8, please. Now, at 3 page 8, therefore, under this document, described as 4 being "Restricted -- policy and commercial" at the top 5 there, if we can highlight, please, the third -- no, 6 sorry, before we do that, let's just read the first 7 couple of lines. It starts with: 8 "the ability of the Post Office to manage changes to 9 the network under each scenario." 10 So they're talking about the various possible ways 11 forward. Then the next bit is "Keys issues will be", 12 okay? Now, the first one is: 13 "how to maximise POCL's existing customer base as 14 benefit payment switches to ACT ..." 15 Fine. Then next one: 16 "how to maintain relations with existing clients who 17 are looking to automation to improve services ..." 18 I'm going to concentrate on the third one. Third 19 bullet point -- and if you could highlight that, 20 Frankie, I'd be very grateful. Thank you: 21 "how to ensure that the subpostmasters (private 22 agents who run the majority of the post office network) 23 perceive that post office business can provide a viable 24 future and do not voluntarily exit the market (reducing 25 the ability of the [Post Office] to manage network 43 1 closures and migrate business to other offices)." 2 Then next paragraph, please, paragraph 21, if we can 3 move just slightly down to that, so it's centred in the 4 page we're looking at -- thank you, and highlight 5 paragraph 21 this time, please. Thank you. 6 Paragraph 21: 7 "Under all options the Post Office will be seeking 8 to manage a reshaping of the network, against 9 a background of commitment to a nationwide network of 10 post offices. Their objective is to retain the current 11 levels of access, especially in rural areas, but to 12 reduce overprovision in some urban and suburban areas, 13 replacing some physical offices with electronic access 14 points. Current trends would in any case see 15 a reduction in the rural network by some 200 offices 16 each year, and a gradual shift to ACT-based methods of 17 payment over time (by 2009/10 almost 50% of claimants 18 are expected to have switched to ACT). Compared to the 19 current network of 19,000 offices, POCL believe that 20 their vision for the future could be served by a network 21 consisting of around 11,000-13,000 full service offices 22 supplemented by 5,000-10,000 electronic access points, 23 many of which could continue to be sited in existing 24 post offices. In practice however we recognise that any 25 network of the future will be constrained by the same 44 1 combination of history and politics that has shaped 2 today's network." 3 So we can see mentions here of the potential impact 4 of Horizon system, we can see that there is reference to 5 ACT matters, to which you're very familiar. 6 Now, can we start, therefore, with the third bullet 7 point: 8 "how to ensure that the subpostmasters ... perceive 9 that post office business can provide viable future ..." 10 Were you aware that the Post Office was seeking to 11 ensure that subpostmasters and mistresses and managers 12 could perceive that the Post Office business could 13 provide a viable future? 14 A. I think we were -- all of us that were associated at 15 a particular level in the programme were well aware that 16 one of the reasons we were doing the Benefit Payment 17 Card was about the future viability of the Post Office 18 network, we knew that. I mean, that was just common 19 knowledge. And even when we came to the decisions about 20 cancelling the Benefit Payment Card, from the work I was 21 doing with ministers and with the Treasury, was looking 22 about how we could smooth the transition to ACT, to help 23 the Post Office network. 24 So we knew there was an impact on the network, there 25 would be an impact on subpostmasters, we knew that, so 45 1 this was about -- by the way, I never, I was not 2 involved in producing this document and I didn't receive 3 it at the time but they were the sort of things that 4 were on our minds. We were aware about the impact on 5 the Post Office network about decisions that were being 6 taken. 7 Q. As it says here, that "subpostmasters (private agents 8 who run the majority of the post office network)", you 9 realise, of course, that subpostmasters are running 10 a small business that they've invested their own time, 11 energy and funds into, yes? 12 A. Yes. 13 Q. You're aware that these are individuals that, no doubt 14 in order to have purchased the physical premises, will 15 have taken out loans and mortgages and the like, yes? 16 A. Yes. 17 Q. You're aware that these are small businesses that depend 18 upon the Post Office income in part and also, of course, 19 on any footfall coming through their post offices as to 20 any other aspects of things that they sell; you're aware 21 of all of that? 22 A. I know of talking to them. I also consulted(?) 23 subpostmasters about matters, yes. 24 Q. Because this appears to be saying to those receiving 25 this "Private and Confidential", commercially marked, 46 1 restricted document, this appears to be saying to 2 everyone that "What we're about to do with the Horizon 3 system is move towards the cutting, by a huge number, of 4 Post Office branches and let's make sure that the 5 subpostmasters and mistresses don't realise that that's 6 about to happen when we're going to implement Horizon". 7 That's what it appears to say, doesn't it? Otherwise 8 the word "perceive": 9 "how to ensure that the subpostmasters ... perceive 10 that post office business can provide a viable future 11 and do not voluntarily exit the market ..." 12 A. You're asking me to offer a personal opinion on this 13 document that I saw this morning, and reading into that 14 literally, it does -- it does look that way. 15 Q. Yeah. 16 A. But from -- again, I think trying to keep this in 17 perspective about what the programme was trying to 18 achieve overall, there were two conflicting -- not two 19 conflicting; there were two sets of requirements here. 20 There was the Department of Social Security's 21 requirements and needs, in terms of payment 22 modernisation, and also the needs of the Post Office 23 network, and I think most people recognised that there 24 was always going to be a tension between those two 25 things. And I think whoever prepared this document, 47 1 I think to some extent, was reflecting the reality of 2 the coming together of those two requirements. 3 Q. Because this seems to be setting out that to achieve 4 a reduction from 19,000 post offices to something like 5 11 to 13,000 offices -- I mean, if we look at that at 6 a rate of some 200 a year, then that's going to take 7 35 years. That can't be right, so this is looking at 8 a reduction to get it within these sort of timescales, 9 being a very fast reduction of -- I can't work it out, 10 but it's well over a third of the post offices that 11 exist. 12 What were you aware of, in terms of making sure that 13 subpostmasters, you know, knew about this intention? 14 A. It was not my role to make sure that subpostmasters -- 15 Q. That wasn't my question. What were you aware of? 16 A. I wasn't aware of what steps the Post Office were taking 17 in discussion with subpostmasters. 18 Q. You see, this appears to show that, by the 19 subpostmasters committing themselves to work with the 20 Post Office on the rollout of Horizon, in a timely 21 fashion, that these were essentially turkeys working 22 their way towards their own Christmas. What do you say 23 about that? 24 A. I think, again, you're asking me to offer a personal 25 view in hindsight, and all I can say is at the time 48 1 these were -- the impact that was going to happen to the 2 Post Office network was entirely a matter for the Post 3 Office network and for their respective ministers, and 4 not the Benefits Agency. 5 We were aware of it and, therefore, there were 6 factors that we took into account in trying to manage 7 the migration to ACT but that was it. 8 Q. Mr Gaskell, this is a briefing to ministers, so this 9 appears to be the Post Office setting out its intention 10 and setting out what it wants to do, "Let's keep it away 11 from subpostmasters"? 12 A. I don't know what was in the minds of the people who 13 were preparing that submission to ministers. 14 MR STEIN: Excuse me one moment. 15 (Pause) 16 Thank you, Mr Gaskell. 17 SIR WYN WILLIAMS: Are there any other questions? 18 MS KENNEDY: Sir, I'm looking around, and I can't see that 19 anyone else is about to stand up. 20 Do you have any questions? 21 SIR WYN WILLIAMS: Fine. 22 Thank you, Mr Gaskell, for, firstly, making your 23 witness statement and, secondly, coming to answer 24 Ms Kennedy's and Mr Stein's questions. I'm very 25 grateful to you. 49 1 A. Thank you. 2 MS KENNEDY: Chair, could I propose that we take a 20-minute 3 break now before the next witness -- 4 SIR WYN WILLIAMS: Certainly, yes. 5 MS KENNEDY: -- coming back at, shall we say 11.35? 6 SIR WYN WILLIAMS: That's fine by me, thank you very much. 7 MS KENNEDY: Thank you. 8 (11.12 am) 9 (A short break) 10 (11.37 am) 11 MR BLAKE: Sir, can you see and hear us? 12 SIR WYN WILLIAMS: Yes, I can. 13 MR BLAKE: The next witness is Sir Adrian Montague. I 14 should say, we may well be finished before lunch. If 15 we're not, we may go through lunch with your permission, 16 sir. 17 SIR WYN WILLIAMS: I am more than happy to complete 18 Sir Adrian's evidence, provided that can be done 19 reasonably, and then we break for the day. All right? 20 MR BLAKE: Thank you very much. 21 SIR ADRIAN MONTAGUE (sworn) 22 Questioned by MR BLAKE 23 MR BLAKE: Can you give your full name, please. 24 A. Adrian Alastair Montague. 25 Q. Sir Adrian, thank you very much for coming today. As 50 1 you know, I ask questions on behalf of the Inquiry. 2 Do you have in front of you a witness statement? 3 A. I do. 4 Q. Is that dated 16 September of this year? 5 A. Yes, it is. 6 Q. Could I ask you to turn to the final page or page 12. 7 Is that your signature on that page? 8 A. Yes, it is. 9 Q. Is that statement true to the best of your knowledge and 10 belief? 11 A. It is, but I have to make two comments on the statement. 12 Firstly, there's a small factual error in paragraph 10. 13 Q. Shall we bring it up, it's WITN04000100. Thank you. 14 Which paragraph would you like to begin with? 15 A. Paragraph 10. 16 Q. Thank you. 17 A. Roughly halfway through that paragraph, you will see 18 I describe the composition of the panel, including Bill 19 Robins and Alec Wylie, and then I say that, 20 respectively, they were the former head of the Northern 21 Ireland Social Security Agency and the former Director 22 General of Communications and Information Services at 23 the MoD. I'm afraid I transposed their affiliations. 24 So Bill Robins, I think, was the former Director 25 General of Communications and Information Services at 51 1 the MoD and Alec Wylie was the former head of the 2 Northern Ireland Social Security Agency. I apologise 3 for the mistake. 4 Q. No apology required. Thank you very much. There is one 5 other change that you have. Do you want to deal with 6 that now or do you want to deal with that at 7 an appropriate time? 8 A. It concerns paragraph 11 but, as we discussed before, 9 before starting, I think it makes more sense -- more 10 sense in the context of the flow of my evidence -- to 11 take it when we come to paragraph 11. 12 Q. Thank you very much. That witness statement and the 13 exhibits will go into evidence and the matters that I'll 14 deal with today are supplementary to that. But I'm 15 going to start with your background. You trained as 16 a lawyer and became a partner at Linklaters; is that 17 right? 18 A. I did. I did. 19 Q. Your expertise was advising on big projects like the 20 Channel Tunnel? 21 A. Yes. 22 Q. You then moved to Kleinwort Benson as head of project 23 finance and later global head of project finance; is 24 that correct? 25 A. I did. 52 1 Q. Then you joined the Private Finance Initiative Taskforce 2 in the Treasury in 1997 until 2000? 3 A. That's correct. 4 Q. After that you have returned to the private sector, you 5 have been chairman of Friends Provident and then 6 chairman of Aviva; is that correct? 7 A. Among other jobs, yes. 8 Q. I'm going to start today with some very broad questions 9 and they relate to you as somebody with significant 10 commercial experience. Can I ask you: what was your 11 view of how the Post Office was run during your 12 involvement in this matter? 13 A. I'll start with three general comments. 14 Q. Absolutely. 15 A. Firstly, like many other witnesses, I'm very aware that 16 the events we're discussing took place 25 years ago. 17 Before I received the papers from the Inquiry's legal 18 team, frankly I had very little residual knowledge of 19 the events of 1998. Those papers have prompted a lot of 20 recollections and I'm using those recollections as the 21 basis of my evidence today, but I'm also conscious that 22 there are still a lot of gaps in my memory. 23 Secondly, in circumstances like this, there's a very 24 strong temptation to comment in hindsight, partly 25 because of the terrible distress suffered by the 53 1 subpostmasters and subpostmistresses, I have dipped into 2 the testimonies from the first phase of the Inquiry and, 3 to be frank, I find them very distressing. So, speaking 4 to counsel for subpostmasters and subpostmistresses, 5 they have my sympathy. 6 That's the reason why you have to struggle against 7 hindsight, and there is also one comment that we'll come 8 to where I think I was speaking in hindsight. That 9 apart, I have tried to speak from my memory of the facts 10 as they stood at the time. 11 Then, thirdly, in preparing for this evidence today, 12 I have read the evidence of Mr Copping, who you -- who 13 the Inquiry interviewed earlier in Phase 2, and I found 14 some of his evidence unexpected. It's also a little 15 awkward because, in some of my recollections, I do 16 differ from Mr Copping and, therefore, I will try put 17 especially the role of the panel in context as we go 18 through this morning. 19 Q. Thank you very much. 20 Trying as best you can, without the benefit of 21 hindsight, what was your opinion at the time of how the 22 Post Office was run? 23 A. Well, I had, I think, limited engagement with the Post 24 Office, because the panel heard from the Post Office on 25 a number of occasions, I obviously saw the Post Office's 54 1 performance through the documents that we had at the 2 time and, I mean, they appeared to me to be reasonably 3 well organised, struggling, I think, with this 4 procurement. I think both the Post Office and the 5 Benefits Agency in different ways struggled with the 6 procurement. 7 I think it was clear that, even if they had been in 8 accord at the outset, by the time I became involved, the 9 Benefits Agency and the Post Office were in different 10 places, and that made life inconvenient for everyone. 11 Q. When you say they struggled, was that a matter of 12 expertise or an issue to do with technical knowledge or 13 something else? 14 A. I think ... I think it was partly their grip over the 15 procurement and the way in which they were supposed to 16 be overseeing the performance of ICL. I think they 17 had -- they had failed to develop a comprehensive and 18 effective approach towards managing the project. But 19 beyond that, it was quite difficult for me to comment on 20 the rest of the Post Office. 21 Q. What were your views of the level of government 22 involvement in the Post Office at that time? 23 A. It was not apparent to me. I mean, what we saw were 24 a limited number of people from the Post Office and from 25 Post Office Counters. On the different working groups 55 1 there were obviously representatives from the DTI, the 2 Post Office's supervisory Department. I had no 3 ministerial contact with anyone except, I think, 4 briefly, Stephen Byers. 5 Q. Was your view of the Post Office one that was 6 independent of government or not? 7 A. I saw -- I saw no evidence that the Post Office was 8 being constrained by government in its handling of the 9 procurement. 10 Q. How about ICL, what was your level of involvement with 11 ICL and your view as to how that was run as a company? 12 A. Well, again, I spent a lot of time with ICL over the 13 period that the panel was involved. If I can just take 14 a step back and think about the chronology here. For 15 the period from sort of March until the summer break, 16 I was quite intensively involved -- 17 Q. This is 1998? 18 A. It's 1998. I was intensively involved through the panel 19 proceedings and preparation and, as it were, mopping up 20 afterwards. In the autumn between September and just 21 before Christmas, there was intensive engagement between 22 the public sector and the Post Office, and the Benefits 23 Agency and ICL, and I came to know ICL, I think, 24 probably better then than I had during the period of the 25 panel's engagement. Then after Christmas, the baton 56 1 really passed to Steve Robson, who was the man in charge 2 of the final negotiations with ICL. 3 Here it's difficult to avoid hindsight. I mean, the 4 impression I had at the time was of an organisation, 5 frankly, struggling, I think, to see its way through 6 technical solutions, commercially an adept organisation, 7 presentations they made to Government were quite slick. 8 But, overall, you know, they were competent people but 9 I think they were being let down by, again, their 10 organisation on the project. 11 Q. Are you talking about a managerial level, below the 12 managerial level, any particular -- 13 A. It's really only the managerial level I had contact 14 with, and board members of ICL. 15 Q. I'm going to begin with procurement, something that you 16 weren't involved in. The tendering process took place 17 in April and May 1996 under the Conservative Government. 18 Am I right to say you had no involvement at that time? 19 A. You are correct. 20 Q. Did you become aware at some stage that there had been 21 three shortlisted suppliers, Cardlink, IBM and Pathway? 22 A. No, not really. I've obviously become aware of it since 23 in looking at some of the testimony. 24 Q. When you say "since", is that because of the Inquiry 25 rather than knowledge at the time? 57 1 A. Yes. I mean, what the panel -- the panel took the view 2 that we were there to try to help the parties in 3 expressing a view on technical viability, then trying to 4 sketch out a way forward, and I think we'd consciously 5 steered away both from any pre-contractual negotiations, 6 which frankly didn't feature at all, and also from 7 trying to attribute blame. This was a panel focused on 8 trying to find constructive solutions to the 9 difficulties that the parties found themselves in. 10 Q. Were you aware to any extent that, as a technical 11 solution, Pathway was the least preferred bidder at that 12 stage? 13 A. No. 14 Q. Were you aware -- and for the record, they may have been 15 documents that you've seen brought on screen for other 16 witnesses, just for the record, it's POL00031237 and 17 POL00028451 -- that the risks that were identified at 18 that procurement stage relating to Pathway were that it 19 could prove unreliable and had a fragile software 20 system? 21 A. We weren't. Perhaps I need to explain how the panel 22 functioned, because I think that would be helpful. So 23 the panel was created in March and April 1998. 24 Bill Robins and I, I think, were strangers to Horizon at 25 that stage. We'd had no previous contact with it. 58 1 I believe I'm right that Alec Wylie had had a previous 2 contact with it, I've seen him referred to in the papers 3 somewhere, I think as part of the programme delivery 4 board, or some such. And it's evidently an extremely 5 complicated procurement. 6 I think it had been badly negotiated, if I'm honest. 7 We had seen, in the taskforce, a number of IT 8 procurements. They are among the most difficult of the 9 PFI projects, because they require a very crisp and 10 complete definition of the sponsor's requirements. PFI, 11 after all, is supposed to concentrate on outputs and 12 outcomes rather than anything very specific in terms of 13 technical specification. But the contractor in this 14 project has to understand precisely what the software 15 is, the hardware and the software is that it's trying to 16 connect to, so ICL would have needed a very precise 17 understanding both of the Benefits Agency's systems and 18 of POCL's systems. 19 And because, you know, the balance of commercial 20 advantage switches from the procurer to the contractor, 21 once the contract is signed, it behoves the procuring 22 authorities to be absolutely specific and categoric in 23 what they're seeking from the contractor, and we know 24 that there were big gaps in some of the contractual 25 documentation. So acceptance testing, model testing, 59 1 none of these things were specified in detail, and so, 2 I mean, almost the two procuring authorities went into 3 this contract with their hands tied behind their backs. 4 There were big gaps which made it very difficult to 5 control ICL and to push the contract to an easy 6 completion. 7 I'm not sure it would ever have an easy completion, 8 but it was a very difficult commercial situation. 9 Q. You say "badly negotiated"; who by? 10 A. Well, by the procuring authorities. You know, they 11 initiated the procurement, it was their requirements 12 that the contractor was responding to. When you go into 13 a procurement like this, as I say, your requirements 14 need to be very, very specific, because that way the 15 contractor knows what he's going to have to deliver, and 16 you've got a decent chance of holding him to account. 17 If a lot is left undefined, it's much, much more 18 difficult. 19 Q. One aspect of the procurement exercise that is highly 20 relevant to PFI is that Pathway was close to the risk 21 transfer sort and would secure PFI clearance but the 22 other bidders wouldn't. 23 Can you explain for us why this risk transfer is 24 relevant and important in PFI? 25 A. I mean, PFI was, at that stage, still a novel way of 60 1 procuring goods and services for the public sector. Its 2 philosophy was that the public sector needed to define 3 its requirements specifically, then the risk of meeting 4 those requirements was transferred to the contractor. 5 As a result of that risk transfer, the contractor would 6 usually have a great deal of authority to define his own 7 methods of working, against a specification dictated 8 from the public sector side, and payment would only 9 usually start once the goods or services had been 10 successfully delivered. 11 So that -- I mean, at one stage in his evidence 12 Mr Copping contrasts PFI with a build and supply 13 arrangement, in which you would expect the procuring 14 authorities to have a much greater interaction with the 15 contractor, more control over the detailed stages of the 16 procurement. But, in this case, in the PFI case, much 17 of that will be left to the contractor, which simply 18 underlines the importance of a precise definition of the 19 procuring authority's requirements at the outset. And 20 from the contractor's perspective, his incentive is to 21 get to the end of the construction period as quickly as 22 possible, because that's the point at which he starts to 23 receive payment. 24 So there is a great difference in risk profile 25 between a PFI procurement and a more traditional build 61 1 and supply arrangement. 2 Q. Looking at this particular exercise -- we'll look at PFI 3 in slightly more detail shortly -- but is there a risk 4 that the party that takes on the highest level of risk 5 will be the one that's selected rather than perhaps the 6 best party for the job? 7 A. It should not work that way, because the procuring 8 authority's appraisal of the competing bids should focus 9 on the level of risk transfer, and you'd expect the 10 procuring authorities or their advisers to comment 11 adversely if one of the contractors was too, if you 12 like, too gung-ho regarding this transfer. I don't know 13 whether it was the case in this situation. 14 Q. Did you have any views at the time about whether Pathway 15 might have been chosen because it was -- 16 A. I had no views. 17 Q. No. 18 I'm going to move on to the Horizon Project Review 19 Group, and can we look at BEIS0000104, please. This was 20 the first meeting of the Horizon Project Review Group. 21 Can you tell us, looking at those who were present, we 22 see names from HMT, DTI, DSS, were those the three 23 Government Departments that formed that group? 24 A. I believe so, yes. 25 Q. How is it that you were selected for that group? 62 1 A. I think perhaps you ought to ask Steve Robson that 2 question, because he was the man who -- he or Harry 3 Bush, perhaps Harry Bush, asked me to attend this 4 meeting. I think it was because the taskforce, my 5 taskforce, had a mandate primarily to focus on new 6 projects, projects pre-financial close, but also they 7 were asked to interest themselves in projects in 8 difficulty, and this was clearly a project in 9 difficulty. And I think it was probably Harry Bush who 10 asked me to attend. 11 Q. Who was Harry Bush? 12 A. He's an official in the Treasury who worked to Steve 13 Robson and was responsible, I think, for this project in 14 the first instance. 15 Q. Thank you very much. Can we scroll down to the section 16 under "The Review Programme", please. This describes 17 two separate stages that the review group would pursue. 18 The first is a "Project Assessment", and it says there: 19 "The first stage would be to determine the viability 20 and potential costs of continuing with the Horizon 21 project. The assessment would need to determine whether 22 Horizon could be delivered, when it could be delivered, 23 what the total costs of delivery would be and the level 24 of risk associated with these assessments. The Benefits 25 Agency, POCL and ICL would need to be involved in this 63 1 stage." 2 The second stage: "Contingency planning for 3 cancellation". 4 Can we go over the page, please, and look at 5 paragraphs 6 and 7. I'm going to read those for the 6 purposes of the record. At paragraph 6, it says: 7 "One approach to this work would have been to 8 commission a firm of external consultants to perform the 9 complete review. But because a decision was needed 10 quickly it seemed probable that PA, who had produced the 11 last report on Horizon, would be the only consultants 12 able to undertake the work. However, there was 13 a significant risk that because of the subsequent work 14 that they had done for other parties involved in the 15 Horizon project, PA would not be in a position to make 16 the judgement required. 17 "After discussion it was agreed that the best 18 solution would be to appoint an assessment board, under 19 an independent chair, to conduct the project review. 20 The board would interview all the interested parties, 21 commission any further work it thought necessary, and 22 then report back (with recommendations). If necessary 23 the panel could use external consultants to do more 24 detailed research under their direction. It was likely 25 that PA might be in a position to do this sort of work 64 1 without compromising their position." 2 So, starting with paragraph 6, it seems as though 3 this task needed to be done quickly. Is it your view 4 that there was limited time in which to undertake this 5 project? Was it sufficient time? 6 A. Erm ... I think the time requirement dictated the 7 panel's approach to the mandate it was given, but 8 within -- within that constraint and with the support of 9 PA, I hope we did a satisfactory job. 10 Q. Did you view it as a quick snapshot, a deep dive or 11 something else? 12 A. No. I think it's ... well, we were given the mandate to 13 consider the technical viability of the project. As 14 I think I said earlier on, two of us on the panel were 15 strangers to Horizon. Doing full justice to that, with 16 the members of the panel directly engaged in inquiry, 17 would have taken a very great deal of time. I don't 18 think any of the members of the panel were free enough 19 from other obligations to devote that much time to the 20 inquiry. 21 So what the panel decided was that it needed to use 22 PA, really, as its devil, to go and make enquiries on 23 the panel's behalf, and I think it's worth just 24 recalling why it was that PA were the obvious people to 25 do this work. 65 1 In his testimony, Mr Copping describes their 2 qualifications for doing the previous Horizon review, 3 which concluded at the end of 1997. I don't think it's 4 necessary, unless you wish to, to bring up the evidence. 5 It's on pages 108 to 110 of Mr Copping's testimony. 6 But when you look at his evidence, it's clear that 7 in that first assignment, in Mr Copping they had 8 a vastly experienced team leader. He said that he'd 9 conducted himself over 30 interviews with the -- I think 10 the ICL staff, perhaps others as well, and there were -- 11 although, for all his experience and expertise in 12 telecommunications and IT, nonetheless there were areas 13 where he needed a team of specialists. 14 So, from the panel's perspective, PA had three 15 powerful recommendations in their favour, you know. 16 They had the expertise, they had the knowledge in depth 17 of the Horizon procurement, and they had resources they 18 could deploy to cover the ground effectively. 19 Now, that's in the context of their review that 20 concluded at the end of 1997, but you can see that from 21 the panel's perspective, the expertise, the resources 22 and the knowledge that PA had, were huge advantages. In 23 fact, I think only through PA could the panel have 24 concluded in the way that they did. PA was the 25 essential tool to allow the panel to do its job. 66 1 Q. Perhaps we could go over the page to paragraph 16, 2 page 3. It says there: 3 "DSS ministers had envisaged the project review 4 being turned around in a period of two to three weeks. 5 The group agreed that this timetable seemed 6 unrealistic -- finding and appointing an external 7 assessors might take two or three weeks." 8 It says over the page: 9 "The timetable would become clearer once the review 10 board had been appointed." 11 From that, it sounds as though there was significant 12 time pressure to complete the job? 13 A. Although that's true, I don't want you to get out of 14 proportion this question of pressure, because, you know, 15 with great respect, ministers often have ideal notions 16 of how long tasks are going to take, and part of the job 17 of the panel was to say, you know, "We need the time we 18 need", and therefore it was two or three months, 19 I think, rather than two or three weeks. 20 Q. At that stage, you considered that PA Consulting was the 21 best for that role? 22 A. As that note envisages, certainly the best, probably the 23 only. 24 Q. Can we look at your report, that is at POL00028094, 25 please. So that's the first page. It was produced in 67 1 July 1998, and you've mentioned those two others who 2 assisted you. Are you aware of any IT experience of the 3 other panel members or indeed yourself? 4 A. Well, starting with myself, I have no training in IT or 5 in engineering. I've never managed an IT project. 6 I mean, I have acquired some experience over the years. 7 I think probably exposure to IT projects rather than 8 experience would be a better way of describing it. 9 Because in the various boards I sat on and as 10 an adviser, we've had to oversee IT projects, for better 11 or for worse, and often for worse. But I have had no 12 direct -- I have no direct qualifications or expertise. 13 Now, Mr Robins and Mr Wylie were appointed to the 14 panel by the Treasury, and all I can say is I developed 15 a high regard for their technical capabilities. I think 16 it's true that Mr Wylie had had a previous connection 17 with the Horizon project, Mr Robins had had none, but 18 their day jobs, I think, involved significant 19 responsibility for the functioning of an IT system and, 20 therefore, I think they were in a good position to 21 complement any expertise that I was able to bring, which 22 lay more in the PFI area than it did in IT. 23 Q. You have suggested that all three of you were also busy 24 doing other things at the time? 25 A. Indeed. 68 1 Q. So although they had experience, were they getting into 2 the weeds or was that somebody else? 3 A. No, I don't think it was our function to get into the 4 weeds. I mean, because of the constraints of time, the 5 panel effectively functioned as a review panel, so we 6 looked to the parties to make submissions to us 7 regarding issues that they wanted to examine -- and 8 perhaps we can come on to this in a minute -- and then 9 PA were mandated by the panel to make the investigations 10 and report back to the panel on their findings. 11 Q. Can we look at the "Executive Summary", that's page 3. 12 I'll start with the third bullet point. It says: 13 "In light concerns over progress, this Panel, 14 chaired by the head of the Treasury Task Force on 15 Private Finance, was set up to make an independent 16 assessment of whether the programme was technically 17 viable, if so how quickly it could be completed and at 18 what cost." 19 Pausing there, technically viable is a term that 20 we've focused on over the past few days; what did you 21 understand "technically viable" to mean? 22 A. Capable in practice of delivering the contracted 23 outputs, but I think the emphasis is on "capable", so we 24 were making a judgement on the ability of the project 25 over time to be developed in a way that would answer the 69 1 specification. 2 Q. Is it similar to feasible, perhaps, or is that something 3 different? 4 A. I'm not sure that I can very precisely distinguish one 5 from the other. 6 Q. Yes. Let's look at the "Findings". If we could scroll 7 down slightly, I'm going to look at the second two 8 bullet points. So it says there: 9 "Our view is that the programme is technically 10 viable. There must be some risk around scalability and 11 robustness because the system has had to be tested at 12 the level of component parts, but we are satisfied these 13 risks are being well managed by Pathway. 14 "There is good evidence of future proofing at all 15 levels. The basic infrastructure is very robust for the 16 future and, in the main, industry standard products have 17 been used. The system should allow POCL to compete for 18 new business in a variety of markets, including banking 19 and financial services. New applications based on 20 smartcard technology should be relatively 21 straightforward and economic. If online applications 22 are required, they may take longer and require more 23 investment." 24 What did you mean there by "The basic infrastructure 25 is very robust" or "robust for the future"? 70 1 A. Well, again, let me go back and explain how the panel 2 and this report were put together. 3 As I said, the -- I think perhaps it might be 4 instructive to go to -- let me see -- I think 5 section 2.4 of annex A to the report. 2.3 and 2.4 is -- 6 Q. Do you have the report in front of you? 7 A. I do. 8 Q. If you were able to give us a page number, that would be 9 very helpful. 10 A. This is 28.04, I think it must be page 30. If I may, 11 I'll perhaps read it out. 12 Q. Absolutely. 13 A. "The first full meeting of the Panel will be attended by 14 all the Parties and the Consultants. The Parties will 15 not be legally represented (either by external or 16 employed lawyers) at this meeting or at any subsequent 17 meetings. At the first meeting the Panel will invite 18 the Parties to make short presentations to the Panel 19 about the outstanding issues and how those issues can be 20 resolved. The Panel will, in its absolute discretion, 21 determine the order in which the presentations are to be 22 given, the number of presentations and the time limits 23 for the presentations. 24 "Following the first meeting the Panel will 25 establish a list of issues to be investigated. It will 71 1 invite further submissions from the Parties and will 2 instruct the Consultants to investigate the matters in 3 dispute. The Consultants will ask the Parties to 4 provide any information the Consultants think could be 5 useful to the investigation." 6 Then: 7 "The Parties shall at all times give such assistance 8 as may be reasonably be requested by the Consultants to 9 enable the investigation to be completed." 10 So there you have, I think, in a nutshell, how the 11 panel intended to work. So, we asked the parties to 12 make presentations to us regarding the issues, we were 13 going to ask PA to go and investigate those issues. 14 Now, I think we will perhaps need to go into this 15 next point in some detail. PA gave us a report, and 16 that report formed the basis of the panel's own report. 17 So many of the judgements in this report were adopted by 18 the panel, but originated in PA's own investigations. 19 Q. Can I just pause you there for one moment? 20 A. Yes. 21 Q. I think Peter Copping's evidence was that there was no 22 separate report and that this report was effectively 23 taking on board the points they made. Am I right in 24 saying that there was a different -- there was, in fact, 25 a separate report from PA? 72 1 A. Well, I think this is obviously an important matter. 2 Now, I think in his witness statement Peter Copping 3 acknowledges that it was likely he submitted to the 4 taskforce some working papers, but he couldn't recall 5 what they were. I don't think he was ever asked whether 6 he had made a report as such and I think that's 7 understandable, because no document has survived, and 8 the panel's report has come to take a very high profile 9 in these proceedings. 10 Now, on this point, I do want to be clear that there 11 was a document from PA, whether it was working papers or 12 a report -- perhaps there's a slight nuance in this 13 case -- but there was a document from PA which laid out 14 in terms the bulk of what appeared in the panel's own 15 report. 16 The panel obviously met to discuss the PA report, 17 and when the panel report came to be prepared, it 18 incorporated the vast bulk of PA's technical advice. 19 Now, I can be quite precise on that because 20 I produced the first draft of the panel report, and 21 I recall taking the document we had received from PA and 22 making minor editorial changes to it. I mean, there 23 were definitions that needed to be changed, nomenclature 24 of the parties needed to be harmonised and I made some 25 changes to the order to bring out the sense, the 73 1 powerfulness of PA's conclusions. 2 But the backbone of this report was PA's own advice 3 to the panel. 4 Q. If we go back to page 3, the findings there, that it's 5 technically viable, is that your language or is that 6 PA's language? 7 A. Well, I think that is probably my language, because 8 I think I wrote the conclusions, but it was on the basis 9 of the later paragraphs in the panel's report which were 10 adopted from PA's own findings given to us. 11 Q. The finding that the programme is "technically viable", 12 and then in the next bullet point that "the basic 13 infrastructure is very robust for the future", is there 14 an intentionally different form of words used in those 15 two bullet points? 16 A. I think if you go further into the report -- I mean, 17 this is essentially a summary of the findings of the 18 report. If you go further into the report, you'll see 19 on pages 11 and 12 -- you may want to go there, but it's 20 helpful to look at it now. 21 Q. Yes, perhaps paragraph 22 might be the starting point. 22 A. Very good. It's -- to go back to the provenance of this 23 report, as I said, the members of the panel were 24 slightly distant from the detailed investigations, 25 because it was PA that carried out those investigations, 74 1 and so these detailed points would have required a very 2 precise understanding of the panel -- of the project. 3 You know, it's not something -- these are not judgements 4 you can make without very close interaction with the 5 different contracting parties in a way that the panel 6 could not itself have achieved because of the time 7 constraints. 8 So that these are judgements that came from PA, we 9 discussed them with PA, we saw no reason to depart from 10 those judgements and because there was a concern to have 11 a panel report rather than another report from PA, as 12 you've already highlighted, we adopted those conclusions 13 for the purposes of our panel report. 14 Now -- so if you look at most of the text of the 15 report, from memory now, I think from paragraphs 14 to 16 99, these were drawn from PA's findings. As I said, 17 I prepared the front few pages, the introduction, and 18 the summary of findings, and annex A, which deals with 19 the way forward. I had also -- I did prepare that, 20 again including some findings from PA. But the text of 21 the report broadly comes from PA. 22 Q. So if we go over the page and look at, for example, 23 paragraph 25, that says: 24 "The main architectural issues are scalability and 25 robustness." 75 1 Or the paragraph below, the final sentence: 2 "We therefore assess the risk of the entire solution 3 failing to operate as expected to be as low as could be 4 achieved in the circumstances." 5 And the following paragraph, which says, 6 for example, that: 7 "... there is a concern that the system is 8 (necessarily) heavily dependent on the third party 9 middleware product 'Riposte'." 10 Are those kinds of findings ones that were made by 11 PA rather than yourself? 12 A. Yes. 13 Q. Would you have scrutinised PA's report in that respect? 14 A. Yes, obviously. So we had this report from PA, there 15 would have been a dialogue between me, in the first 16 instance, and then the panel around the observations and 17 the conclusions, and then, as I said, I would have 18 prepared -- I did prepare -- the first draft of the 19 panel report, and that would have been circulated to the 20 other members of the panel for their comments, and then 21 sent to the sponsors, the inter-ministerial committee. 22 Q. Those kinds of issues that are highlighted on this page 23 that we see now, were they a cause for concern at all at 24 the time? 25 A. Erm ... no, I think we'll come to this. I think that 76 1 the prevailing mood, I think, in the sponsors and in PA 2 and in the panel, was of concern at the way the 3 procurement was progressing, but I think none of the 4 very severe dysfunctionality that came to dog Horizon in 5 its later years, that was not apparent, I think, to any 6 of us. 7 If you remember, I said that the methodology of the 8 panel was that we relied on the sponsors to make 9 presentations to us. We would have created this list of 10 issues and then we remitted those lists of issues to PA 11 for detailed investigation. And so, I think, there was 12 concern as to the way that the procurement was 13 progressing, but none of the technical issues that came 14 to dog the project were apparent to us or brought to our 15 attention, I mean, either by the parties or by PA. 16 Q. Did you think at the time that PA were carrying out 17 a thorough enough investigation? 18 A. Well, I mean, we were slightly comforted and reassured 19 by the fact that PA had conducted a very intensive 20 investigation into the project a few months previously, 21 and, I mean, I think my expectation was that Mr Copping 22 would have refreshed those conclusions in dialogue with 23 the sponsors. I mean, I think in his evidence quite 24 a lot was made of the fact that he was not specifically 25 asked to make any enquiries. To be honest, I think 77 1 I find it difficult to believe he did not make any 2 enquiries, because he knew the project so well. He was 3 going to give the panel a report on technical viability, 4 and, therefore, I think he would have at least needed to 5 have refreshed his discussions with the sponsors and 6 with ICL. 7 Q. What is your view of the evidence that he has given to 8 the Inquiry in respect of the production of this report? 9 A. Well, I don't think he was specifically asked whether he 10 had produced this report or the report on which the 11 panel's findings were based. You know, it's 25 years 12 ago. I think it's perhaps a case that recollections may 13 differ. I mean, he's a very experienced consultant. 14 It's many years since I've spoken to him but he was very 15 professional, and I think he did do a good job for the 16 panel. 17 Q. You say recollections differ; how so, in relation to 18 this particular event? 19 A. If you recall his evidence, I mean, he was asked 20 specifically a number of questions regarding his 21 interaction with ICL and BA and POCL during the process 22 of his investigations, and he said, I think in answer to 23 all of them, that he made -- he had not asked any 24 questions. He's a very experienced consultant, he has 25 a mandate from the panel to produce a report, 78 1 affecting -- regarding technical viability; personally 2 I'd have thought it was likely he had a dialogue with 3 the -- in fact, I think at one stage he does say he had 4 a number of meetings with the project participants. 5 I think there must have been a flow of questions to and 6 fro. 7 Q. I'm going to ask you about the use of your report. Can 8 we look at BEIS0000418, please. I don't know if you saw 9 Mr Sibbick's evidence at all, from earlier -- 10 A. Briefly. Briefly. 11 Q. This is a letter from Peter Mandelson, Secretary of 12 State for Trade and Industry at that time, to 13 Stephen Byers, who was the chief secretary to the 14 Treasury. 15 If we go over the page, please, he says there: 16 "There is still some way to go to complete the 17 Horizon project, but the basic development work has been 18 thoroughly evaluated by independent experts who have 19 pronounced it viable, robust and of a design which 20 should accommodate future technological developments." 21 Then in the next paragraph it says: 22 "I believe the only sensible choice is to proceed 23 with the Horizon project. It is the way forward which 24 offers the least commercial and technological risk." 25 Starting with that first paragraph there, if that is 79 1 a description of your report, do you think that that is 2 an accurate summary of the findings that you made: 3 "viable, robust and of a design which should accommodate 4 future technological developments"? 5 A. I think what the report -- the panel report said was 6 that we thought that the project was "technically 7 viable", I don't think we specifically comment on its 8 general robustness. And we said it had been future 9 proofed. Now, that may be consonant with a description 10 that the design should accommodate future technological 11 developments, but I think this is a minister writing to 12 a colleague, and I think some licence with the 13 conclusions is only to be expected. 14 Q. As an Inquiry, we're interested in how this phrase 15 "robust" keeps on cropping up over the years. Is it 16 your evidence that your report did not pronounce it as 17 "robust"? 18 A. Yes. 19 Q. In relation to technological risk, did your report say 20 that it offers the least technological risk? 21 A. I don't think so. I don't recognise those words. If 22 you can point me to something that says the opposite, 23 I'll consider it, but I don't think so. 24 Q. In relation to the reference to independent experts, are 25 you aware of any other independent experts having 80 1 pronounced anything in that kind of a period, so summer 2 to December 1998, or is it likely that that is 3 a reference to your report? 4 A. I think the chronology suggests it is a reference to my 5 report. There was subsequently a review by, I think, 6 KPMG at Mr Corbett's behest, which came to conclusions 7 on all fours with the conclusions that the panel came 8 to. 9 SIR WYN WILLIAMS: Sir Adrian, Mr Blake has told you we're 10 trying to find the origin of the word "robust" in the 11 context of Horizon. This may be an impossible question, 12 and please tell me if it is, but do you think that the 13 word "robust" in your report is a word you would have 14 chosen or a word you would have adopted? 15 A. I think it's probably a word that I would have adopted, 16 sir, but, you know, if we take a step back for a minute, 17 what I think we saw was a procurement in -- if I'm 18 honest, in disarray and difficulty. I've read some of 19 the technical evidence submitted to the panel, and 20 I find the lack of professionalism in ICL quite 21 disturbing. I think it was not obvious to us that that 22 was the case. I think ICL had a tendency to play its 23 cards quite close to its chest and I think it was 24 feeling slightly defensive about its ability to see this 25 through to the end. 81 1 But, again, I'd say that the defects that became 2 apparent later in the process were not obvious to people 3 at the time we made the report, certainly were not 4 obvious to the panel, and I think, if they had been 5 obvious to the parties, in the list of issues they were 6 going to prepare for the panel, I can't see any reason 7 why they would not have mentioned it. 8 And I believe that PA itself had no inkling of these 9 issues, so that they may have been -- they may have been 10 brewing in the undergrowth, but none of them were 11 apparent at this stage in the process. 12 MR BLAKE: While we're on that, perhaps I can take you to 13 FUJ00080690. This isn't a document that was in your 14 pack before the hearing today. It's not a document you 15 would have seen at all at the time, and it may be 16 something that you've seen following this Inquiry, and 17 it's a report of something called the EPOSS PinICL 18 Taskforce at ICL. It was a taskforce that, as it says 19 there, took place between 19 August and 18 September 20 1998, and a PinICL is effectively an incident report 21 within ICL. 22 Can we very briefly look at page 4, please. In the 23 "Introduction" there, this should give you enough 24 flavour. I don't know if this is something you have 25 seen mentioned at all during this Inquiry or -- 82 1 A. No, I have not seen this. 2 Q. No. I'll take you to it very briefly and, please, if 3 you want to spend more time on it, I'm happy to, but 4 I don't think I need to for the purposes of the 5 question. 6 It says about halfway down that second paragraph: 7 "During the course of the Task Force it became clear 8 that there are significant deficiencies in the EPOSS 9 product, its code and design, and these are also 10 presented in this report." 11 If we go down to the "Management Summary" -- these 12 are just examples to give you a flavour of this 13 report -- it says: 14 "Before the EPOSS Task Force was initiated [so 15 that's pre-August 1998] the Counter Development Team 16 were immersed in a seemingly impossible task of dealing 17 with PinICLs [so incident reports] that were being 18 raised faster than they could be cleared." 19 Then perhaps we can go to page 7. There is 20 a section on the "EPOSS Code" that we've looked at in 21 quite a lot of detail in this Inquiry: 22 "It is clear that senior members of the Task Force 23 are extremely concerned about the quality of code in the 24 EPOSS product", et cetera, et cetera. 25 Were these kinds of concerns brought to your 83 1 attention in the summer of 1998? 2 A. No, I think, to be clear, the taskforce here is 3 a different taskforce from the Treasury taskforce -- 4 Q. Oh, absolutely. 5 A. That's just worth clarifying. 6 Q. Yes. 7 A. The answer is no. No, these were not brought to our 8 attention. I mean, I think again it's worth a comment, 9 because I think in processes of developing, you know, 10 software and detailed design, incidents will occur and, 11 you know, they have to be dealt with in order of 12 priority, and there's clearly an avalanche of incidents 13 for ICL here. And I think it perhaps takes time before 14 how they interact together becomes apparent. 15 Now, this was the sort of information which it would 16 have been very valuable for the taskforce -- sorry, for 17 the panel to have, and it is the sort of information 18 which one would have hoped, had it been available at the 19 time, the parties brought to the panel in the list of 20 issues to be considered, but it was not the case. 21 Q. Thank you. Perhaps we can look at page 18 of that 22 report to give you a little more flavour. It says: 23 "Whoever wrote this code clearly has no 24 understanding of elementary mathematics or the most 25 basic rules of programming." 84 1 That's in reference to one example of code. 2 Can we go back to your report at POL00028094 and 3 look at page 32. This is a list of meetings that you 4 had with ICL and POCL. So am I right to say that, as 5 your panel, you sat -- and I think you've referred today 6 already to presentations that were made. So on 11 May 7 there was the initial presentation by all parties; there 8 were then private presentations of issues by parties on 9 11 May; 19 May, presentations of parties' business 10 cases; 28 May, demonstration to the panel of Horizon 11 products; 8 June, final presentations by the parties; 12 and then, 11 June, presentation to parties of panel's 13 initial findings. 14 Did ICL give you any inkling, at this time -- so 15 quite close in proximity to that taskforce -- of those 16 kinds of issues that they were experiencing? 17 A. No, they didn't. 18 Q. Should they have? 19 A. I think it depends on their state of knowledge regarding 20 these incidents. I mean, I note that the report you 21 referred to a moment ago was dated in September. This 22 was five months previously. Things probably developed 23 quite quickly. If there had been, I think, 24 a significant possibility that the code could be 25 dysfunctional, you would certainly have hoped that, in 85 1 all candour, they would have mentioned it. And you 2 would have -- if the parties had had that inkling, you'd 3 have expected them to mention it. I mean, I think the 4 difficulty is the panel -- the panel was a recipient of 5 all these advices from the parties and our job, with the 6 assistance of PA, was to try to assess them overall. 7 I think there was probably an element of 8 salesmanship, defensive positioning on ICL's part. 9 I don't recall the precise interactions between us and 10 the parties in those sessions that you referred to. 11 Unfortunately, I don't think any of the documents have 12 survived. But I expect they were, you know, reasonably 13 detailed presentations. But I think -- in ICL's case, 14 I think it will be natural that they put slightly 15 a positive spin on events. 16 Q. Were PA Consulting tasked with finding out, under the 17 spin, the truth of the matter? 18 A. Well, you know, we asked them to give us material that 19 we could use in the panel report in the areas that 20 they've covered in their own report and, you know, they 21 were tasked with advising the panel regarding the 22 technical viability of the project. I think -- I think 23 they were themselves, you know, looking at it with the 24 benefit of the detailed report they'd done the previous 25 year. As I've said, I think that, in order to complete 86 1 the report that we had, they'd have needed to interact, 2 to use a neutral word, with ICL, and they were much 3 closer to the detail than the panel could possibly have 4 been. 5 But, equally, you know, they're not all seeing, all 6 knowing either, so they're rather dependent on what 7 people tell them. I mean, they could have and clearly 8 they did make detailed enquiries because, otherwise, you 9 know, the detailed list of recommendations could not 10 have been produced by the panel. So they've obviously 11 had an interaction with Pathway. And, as I say, I don't 12 think -- I don't think it could have been disclosed at 13 this stage, (inaudible) at this stage. 14 Q. I want to ask you about further monitoring after your 15 report was produced. Can we look at HMT00000021, 16 please. Thank you very much. This is a note from you 17 to Harry Bush, Harry Bush being in the Treasury? 18 A. Harry Bush being someone who reported to Steve Robson 19 who was, I think, in charge of this project and who was 20 the person who had invited me to attend that first 21 meeting in March. 22 Q. Thank you. 23 So this is 6 July 1998, and your report was produced 24 in July 1998, this came after your report was provided? 25 A. Certainly after the initial findings had been 87 1 communicated to the parties. I've seen somewhere 2 a suggestion that my -- the panel report is not 3 available until 22 July. It was some time in July. 4 There may have been a draft available at this point. 5 Q. Thank you. 6 I'll just read a little bit from paragraph 1, it 7 says: 8 "This minute addresses three issues what we should 9 be doing to try to stabilise the Horizon programme in 10 the short term to prevent any further deterioration 11 pending Ministers' decisions, whether there is any 12 compromise between Option 1 and Option 2 which might 13 command the support of both BA and POCL, and who the 14 troubleshooter might be." 15 We have heard a lot about this period, I don't want 16 to spend any time on it really, can you briefly 17 summarise for us what was going on at a political level 18 in July 1998 with regards to various options? 19 A. I think the political level was not visible to me. 20 I mean, I think that the background to this note was 21 that I think we must have had a meeting with POCL and 22 the Benefits Agency as a follow-up to the meeting when 23 we told them about our initial findings. I think it's 24 clear -- and to be clear, you know, the way forward that 25 we suggested was going to involve some pain on the part 88 1 of all three of the parties to the transaction. 2 And what we were trying to do was to find a way 3 forward that was -- kept the greatest value for money 4 and involved the least disturbance, and so we did opt 5 for what's known as option 1, the continuation of the 6 project with some changes. And I think we must have 7 discussed this with the Benefits Agency and, not to put 8 too fine a point on it, they were having none of it. 9 You know, we had got to the stage where the Benefits 10 Agency were far removed from their -- the support they 11 must have given the project at the outset, because it 12 was a joint procurement, and had moved to the point 13 where they had a strong preference for ACT, Automatic 14 Credit Transfer, and they weren't going to willingly 15 compromise. 16 So I think, you know, after that meeting, I was 17 really rehearsing with Harry Bush whether there was any 18 alternative that might allow us to cut through this 19 logjam on the public sector side. 20 Q. Can we look at the first bullet point, please, if we 21 could scroll down slightly. I'm going to read what it 22 says there for the purpose of the record, it says: 23 "The Panel report has a list of actions which need 24 to be completed by the end of July. Some of these 25 presuppose ministerial decisions one way or the other, 89 1 and will therefore have to be put to one side for the 2 time being (essentially those affecting the 'strategic 3 issues'), but many of the actions relating to the 4 critical or operational issues will adversely impact the 5 future timing of the programme if they are not pushed 6 through on a timely basis. We need to frogmarch the 7 parties into resolving these issues. To encourage them 8 to do so, we discussed a weekly forum to monitor 9 progress. This might meet [for want of a better 10 alternative] under our chairmanship, and might be 11 informed by a hands-on monitoring role on the part of 12 PA. I have spoken to Peter Copping about this and he 13 has sent in a proposal, which I attach. The price is 14 significant, and we need to consider whether the role 15 represents value for money." 16 So you're there proposing a hands-on role by 17 a technical expert, is it? 18 A. Erm, I think it would need -- it would need to have had 19 direct official input to command the right degree of 20 authority, but clearly a technical expert would have 21 been invaluable to help the chairman, whoever it was 22 going to be, conduct the forum effectively. 23 Q. Do you know if that role actually was established? 24 A. I don't know. I mean, I think I was a bit outside my 25 brief in making this suggestion. You know, I think with 90 1 the conclusion the panel report, the leadership in the 2 discussions switched back to the officials in all three 3 Departments who had had the conduct of it prior to the 4 panel, and I think this was slightly outside my brief. 5 Q. It may have been outside your brief but it looks there 6 as though you are suggesting that there needs to be more 7 follow-up once your report has concluded; is that a fair 8 summary of that paragraph? 9 A. Well, it is indeed a fair summary. You know, I think 10 that the Treasury's way of working tends to sit over the 11 Departments and check that the Departments are handling 12 things effectively. It's quite rare for the Treasury to 13 take the lead in assuming control over the process, and 14 I think I was perhaps outside my brief in assuming that 15 that could be done. 16 Q. When you wrote your report, did you see that as 17 a snapshot in time, or did you see it as something that 18 could be used for the next six months, a year -- 19 A. I think it was inevitably a snapshot in time, because, 20 you know, the fundamental tenet of that report was that 21 it interacted with the submission of the parties and the 22 report of PA, and things were moving evidently quite 23 fast and, therefore, I think it would need to have been 24 refreshed over the months that followed. 25 Q. Do you think that that was clear to those who were 91 1 involved in the project? 2 A. I honestly can't tell you that. I mean, I think one of 3 the challenges of the public sector is that, sometimes, 4 these reports occupy a terribly important part of the 5 debate and you don't continually ask yourself how the 6 situation has evolved. So -- but I think it did come to 7 have a slightly totemic value. 8 Q. Can we look at page 4 of this note, please, and it's the 9 final paragraph. Thank you. 10 We have heard a little bit about the role of 11 Graham Corbett. Can you tell us a little bit about that 12 and how he was chosen? 13 A. Well, I think recognising that any of the options we had 14 considered would require some difficult decisions to be 15 taken by all the parties, I thought we believed that the 16 intervention of a third-party troubleshooter with 17 a mandate to try to provide an acceptable way forward 18 for all parties -- or an acceptable way forward, not 19 an ideal way forward -- would be a useful contribution. 20 And so the troubleshooter was recommended, suggested in 21 the panel report and then the question was: who should 22 the troubleshooter be? 23 As I said in this note, Steve Robson came up with 24 the idea of Graham Corbett. I had known Graham Corbett 25 quite well in the past, because he had been the finance 92 1 director of -- CFO of Eurotunnel, and he's a very robust 2 character -- he was a very robust character, so 3 I thought in terms of someone who'd be able to immerse 4 himself in the detail, financial details in particular, 5 of the procurement at the time that it's -- it was -- 6 you know, at this time. I thought he would have, 7 potentially have a very positive influence on the 8 outcome. 9 Q. You have said that he may be able to "knock heads 10 together if the parties are unable to hammer out 11 a detailed commercial solution for themselves". I think 12 you also noted that he wasn't, in brackets, about 13 halfway down that paragraph, he's not, as far as you 14 know, in the IT field. Was his report intended to be 15 a look into the IT side? 16 A. Not so far as I'm aware. I didn't see his terms of 17 reference. Indeed, I wasn't connected -- I wasn't 18 concerned with his report. I do not believe he would 19 have held himself out as having any capability to 20 address technical details, but he's fundamentally 21 a financial and a commercial man, and his role could 22 have been very decisive, I think. 23 Q. Thank you very much. 24 Can we look at POL00028098, please. Perhaps we 25 could go to page 3. This is his report. Did you see 93 1 his report at the time? 2 A. I can't remember. I may have done. 3 Q. Okay. I'll only look at it very briefly. It starts 4 there with "Background", it says: 5 "On 17 September [this is to the chief secretary to 6 the Treasury] you appointed me as independent adviser to 7 this project with the terms of reference set out in 8 annex A ... and B [and something] with a reporting date 9 of 16 October." 10 So, again, this is a very short timescale for 11 a report, similar to yours in many ways, in terms of the 12 timescale? 13 A. In terms of timescale, I think he was under probably 14 even greater pressure than the panel was to come up with 15 some findings. I think it has a slightly different 16 focus, because, as he says in that first paragraph, he 17 had been unable to reach agreement on a commercial basis 18 for proceeding, so I think that was clearly the focus of 19 his efforts. 20 Q. Can we look at page 6, please. At the bottom of that 21 page, it sets out another role for PA Consulting. It 22 says: 23 "Soon after my appointment we established a working 24 group of the programme directors from each of the 25 parties working under the chairmanship of the Director 94 1 of the Horizon Programme Office ... now situated within 2 POCL, and with PA Consulting keeping close to and 3 guiding their discussions. PA's final report as of 4 14 October is attached at Annex E together with the 5 HPO's high level end-to-end plan for the programme which 6 has been signed off by all parties. Attention is drawn 7 to the generally encouraging overall summary at the end 8 of the PA report, but also to the continuing high risk 9 area of acceptance procedures, both the definition of 10 the tests themselves and the consequences of failure." 11 Perhaps we can look at their report, it's at 12 page 32. I don't need to go into this in any detail, 13 but perhaps we can look at this very briefly and also 14 over the page. 15 Were you aware, at that stage, that PA had carried 16 out this work? 17 A. Erm ... I don't have any very clear memory of it. I may 18 have been. 19 Q. It looks there to be quite a high level analysis. Could 20 we look at the page before, please? Thank you very 21 much. Did you have in mind something more significant 22 than this kind of work when you referred to hands-on 23 monitoring? 24 A. Yes. I mean, I've had quite a lot of experience with 25 troubled -- projects in difficulty and, generally 95 1 speaking, you know, the parties are a long way apart 2 commercially. You know, there may be unclear technical 3 issues, and, I mean, I think it's characteristic of 4 those projects, shared by this project, that there's 5 a sort of denial, a refusal to get down to basics and 6 understand precisely what the issues are. 7 And I think what you need is what Graham Corbett 8 provided on a commercial basis for a short period of 9 time, which is a heavyweight who takes the chair, forces 10 people to confront the issues, and does bang heads 11 together. That's the way you rescue a project. 12 I didn't see any of that, well, at any time really, in 13 the autumn of 1998. That's what was required. That's 14 essentially what the panel had suggested. 15 Q. Thank you. 16 Can we look at POL00031114, please. Now, you -- 17 well, I'll ask you actually. Are you aware of a further 18 report that was carried out by Project Mentors? Perhaps 19 we could go over the page. This is a summary of the 20 report, and it says there: 21 "As you will see, all three of Andrew's team [that's 22 Project Mentors] are (I quote from Andrew's letter to 23 me) 'deeply concerned that their findings show a serious 24 problem with the way in which ICL Pathway have developed 25 the system. The impact of this is likely to be that 96 1 there will be failures to meet essential user 2 requirements, causing the need for extensive re-work 3 before the system can be accepted and, potentially, 4 operational problems if the system is rolled out." 5 That is eight days after the letter that I showed 6 you earlier from Peter Mandelson, referring to the 7 system as "robust". 8 Were you aware at this stage or at any other stage 9 of a further review being undertaken into Horizon, 10 a further technical analysis such as this? 11 A. I was not aware of the Project -- 12 Q. Mentors. 13 A. -- Mentors report, until it landed on my desk last week. 14 Q. Yes. Were you aware of any other significant report, 15 prior to the rollout of Horizon, into the robustness or 16 reliability of Horizon? 17 A. I was not. But I don't think I could be expected to, 18 because, you know, in the autumn of 1998 my involvement 19 was significantly less than it had been during the 20 summer. 21 Q. To what extent were you involved in late 1998 and into 22 1999 with the Horizon project? 23 A. I noted from some of the papers that I had been asked to 24 act as a commercial facilitator at one stage. This is 25 in December, I think, 1998. And ... I'll go back 97 1 slightly. 2 I think there were four attempts to secure 3 a financial accommodation with ICL during the autumn. 4 Graham Corbett's was the first. There was then a letter 5 from the chief secretary, I believe, to ICL saying 6 "You've got to come up with some sensible proposals". 7 ICL responded. The proposals were not regarded as 8 sensible or acceptable. There was a last ditch 9 negotiation, which I think I was involved in. I have no 10 memory of what happened in that -- at that stage. And 11 then, finally, just before Christmas, probably under 12 extreme pressure, ICL volunteered some further 13 reductions. But these were negotiations really 14 conducted at a high commercial level, and did not 15 involve any recalibration of the technical assessments. 16 I think this is quite an interesting note, again 17 I hadn't seen it, but I think it's a sort of 18 pre-litigation note. It's written under the cover of 19 the legal advisers' correspondence. I'm sure it's 20 accurate. I mean, I think it's perhaps -- viewed from 21 a litigation perspective, you'd expect it, I think, to 22 be quite critical of ICL, it obviously is quite critical 23 of ICL. I think some of the evidence I have seen since, 24 in the earlier phases of this Inquiry, now suggest that 25 the observations may well have been accurate. But, as 98 1 I say, I had not seen that report until last week. 2 MR BLAKE: Thank you. 3 Sir, I have about ten more minutes of questions, so 4 I'm going to continue, unless you feel we need a break. 5 SIR WYN WILLIAMS: Well, as it happens, I would quite like 6 a short comfort break, but I literally mean short, so 7 all wait for me to disappear and then return in a minute 8 or two, if that's all right. 9 MR BLAKE: You may not be the only one. Perhaps if we break 10 for ten minutes, but then we'll sit through lunch, 11 because I won't be very long at all. 12 SIR WYN WILLIAMS: Are there any questions other than yours, 13 Mr Blake? 14 MR BLAKE: I think it's unlikely. Mr Stein looks like he 15 may but, if he does, it will be brief, I'm sure. 16 MR STEIN: Sir, if I can just add, we may have one question 17 but I may be able to discuss that with Mr Blake so that 18 we can deal with it. 19 SIR WYN WILLIAMS: Let's have five minutes for my comfort 20 and everyone else's who needs it, and then we'll carry 21 on. 22 MR BLAKE: Thank you very much. 23 A. Perhaps I may join you, sir. 24 (12.56 pm) 25 (A short break) 99 1 (1.03 pm) 2 MR BLAKE: Thank you very much, sir. Can you see and hear 3 me? 4 SIR WYN WILLIAMS: Yes, I can, thank you. 5 MR BLAKE: Thank you. 6 I'm going to ask you a few questions about PFI. Was 7 this the largest PFI project you dealt with? 8 A. No, there were quite a few larger than this, it was 9 probably the largest IT project. 10 Q. A witness that we've heard from, Mr Folkes from the Post 11 Office, has described PFI as creating a black box where 12 the service provider's job was to ensure it created the 13 right output but the contents of the box weren't 14 available, so how it worked, how it was built. Do you 15 agree with that? 16 A. No, not really. I think I said earlier on that the task 17 of the procuring authorities in a PFI project is to be 18 absolutely specific regarding their requirements. Now, 19 I think the philosophy of PFI is, with a very specific 20 output specification in front of them, it is for the 21 contractor to decide how to achieve that specification. 22 But that does not mean that the procuring authority 23 sort of goes away and comes back again in four years' 24 time. The procuring authority should reserve to itself 25 rights to check progress, to be informed as to the way 100 1 in which the detailed engineering phase is being 2 undertaken, and there should be check points where the 3 contractor has got to produce evidence of testing or 4 evidence of how the specification is being undertaken. 5 So the essential difference is that the contractor 6 has probably a greater influence over its methods of 7 working than you would necessarily expect in 8 a conventional procurement, but it should not mean that 9 the procuring authority has no control of it. 10 There was, as an example, in PFI a very strong 11 procurement, at around the same time, handled by the 12 Treasury taskforce, and it concerned the Department of 13 Treasury, the National Savings. 14 It was a business transformation project, it was not 15 of the same scale as Horizon, but it was very 16 complicated because it involved the transfer to 17 a private sector contractor of the majority of national 18 savings business processes, so a large element of 19 re-computerisation, a large transfer of civil servants 20 into the private sector. 21 This transaction was reviewed by the National Audit 22 Office and was commended as the right approach to 23 transactions of this sort. It's clear that Horizon was, 24 you know, conducted at an earlier phase when there was 25 less understanding of what made for a good PFI project, 101 1 but I think "black box" is too simplistic a description. 2 Q. I think you explained at the beginning of your evidence 3 that you were quite critical of the negotiation process. 4 Did you think at the time -- were you aware of concerns 5 about the sharing of information -- 6 A. No, I think I was -- I mean, you asked me general 7 questions, which didn't specifically focus on my state 8 of knowledge regarding the negotiations but, in view of 9 what I know now, I would have been critical of them 10 because I think they were -- they failed to take 11 advantage of the fact that, in a PFI project, until the 12 contracts are signed, the balance of negotiating 13 advantage is always with the procurer because the 14 contractor wants the business. Afterwards the balance 15 shifts a bit towards the contractor. 16 Q. So is it your view that contractual provisions regarding 17 the sharing of information should have been agreed at 18 an early stage? 19 A. I think the full scope of the intervention that 20 a procuring authority might want to make during the 21 course of production of the platform, production of 22 Horizon, that ought to have been specified. 23 Q. Were you aware at the time of concerns about the sharing 24 of information or -- 25 A. No. 102 1 Q. -- lack of sharing? 2 A. No. 3 Q. There were renegotiations about the contract when the 4 Benefits Agency had pulled out of the project. Was 5 increasing access to information from ICL something that 6 was raised with you at the time or discussed at all? 7 A. I wasn't a part of those negotiations, so I'm afraid 8 I can't answer the question. 9 Q. Do you think that a project of this kind was suited to 10 PFI? 11 A. I think the experience of PFI generally -- I'll come 12 back to the specific question -- the experience of PFI 13 generally is that it required a good deal of 14 sophistication on the part of the procuring authorities. 15 Dealing with PFI was not a job that many civil servants 16 had training in or experience of and, therefore, a very 17 important role was taken by the financial advisers to 18 help them structure a procurement effectively, and there 19 were some procurements where the Treasury taskforce 20 acted as a sort of in-house adviser to the Department 21 concerned, and when you had the right combination of 22 experienced civil servants, the right calibre of 23 financial advice, and sometimes the involvement of the 24 Taskforce, then I think projects could come to fruition 25 very successfully. 103 1 I mean, an example is the Prison Service, which 2 conducted PFI procurements to the same model over five 3 or six different, usually, prisons and, by the time they 4 got to the end of this, they were a cracking good 5 procurement unit because they had all the experience, 6 they had all the contacts, they were very effective. 7 But first time procurements with inexperienced civil 8 servants, perhaps with the wrong calibre of financial 9 advice -- I don't know who gave the financial advice on 10 this project -- and a lack of determination to extract 11 maximum advantage from the opposition as the procurer 12 before signing, that's a recipe for a bad outcome. 13 Q. The Chair will, in due course, be considering 14 recommendations to make to avoid problems in the future. 15 Is there something that you can identify in that regard, 16 with regard to the PFI project, that should be avoided 17 in the future or that should be done better in the 18 future? 19 A. I can think of lots of things. I mean, PFI has rather 20 fallen out of fashion, I think because of indifferent 21 quality of execution of these projects. I think that 22 the well-executed projects are outstanding, the badly 23 executed projects have attracted all the adverse 24 commentary. I think it is a combination of experience 25 in the Civil Service to manage these procurements, the 104 1 right calibre of financial advice, an understanding of 2 the risks you're trying to transfer, and an awareness of 3 how you control this transfer during procurement. 4 I might point you perhaps to the National Savings 5 outsourcing, where the NAO reviewed it in detail and 6 came up with a number of positive recommendations. 7 Q. We discussed earlier that part of the PFI contract was 8 for the contractor to bear the risks, and that, as we've 9 heard, was one of the reasons why ICL won the project in 10 the first place. Were you aware in 1998 of ICL 11 demanding a positive return on their investment and 12 a shift away from that risk-based model? 13 A. Well, the financial negotiations in the autumn of 1998 14 clearly showed that ICL wanted to repair its finances 15 and secure improvements in the financial terms to 16 achieve a positive NPV, and I think the interactions 17 that Graham Corbett had with ICL were quite important 18 because they showed a reluctance on the part of the 19 public sector parties to go as far as ICL was seeking. 20 Q. Do you think that, ultimately, ICL obtained the benefits 21 from PFI, in respect of determining the solution but 22 without ultimately bearing the risks? 23 A. I think you'd have to ask ICL that. I'm afraid I can't 24 comment. 25 Q. Did you have an opinion at the time? 105 1 A. Well, you know, I wasn't surprised that ICL were 2 projecting to suffer a considerable loss on the 3 contract. They had clearly underestimated the risk. 4 I mean, the Project Mentors report shows that they were 5 ill prepared for the size of the task that faced them. 6 They were unprofessional in the way they handled it. 7 None of this was apparent earlier in 1998 and so 8 it's not a surprise that they were going to suffer 9 a loss. It's not a surprise that they were trying to 10 recover it. I think, you know, successive rounds of 11 negotiation were quite robust in denying ICL the extent 12 of the improvements in their finances that they were 13 claiming, but ICL was on the hook for this contract. 14 I can comment until, I think, Christmas 1998. When 15 Steve Robson took over the baton in early 1999, I had -- 16 I was copied in on some of the documents, although 17 I really had no continuing involvement with their 18 discussions. 19 Q. I want to ask you about prosecutions. Did you give any 20 thought as to the Post Office's prosecutions of 21 individuals using data from Horizon during your 22 involvement in the project? 23 A. None at all. I mean, I -- I don't think that we were 24 aware at this stage -- I don't understand if there were 25 prosecutions at this stage but, if there were, we were 106 1 certainly not aware of them. 2 Q. Was the reliability of accounting data something that 3 you were asked to look into as part of your report? 4 A. No. 5 Q. You left the project in the New Year of 1999, and Steve 6 Robson took over, as you've explained. Who was Steve 7 Robson? 8 A. Steve Robson, I think he was the Second Permanent 9 Secretary at the Treasury. 10 Q. Why was he a suitable person to take over? 11 A. Well, I mean, the panel will be taking evidence from him 12 in a few days' time, you'll make your own judgements of 13 his suitability. He is an extraordinary civil servant, 14 very senior, very experienced, very clear in his 15 judgements. I think if anyone had had the opportunity 16 to turn this around, it would have been Steve Robson. 17 Q. Thank you. I don't have any further questions. Is 18 there anything that you would like to say at this stage? 19 A. No, I've already, I think, expressed the sympathy I feel 20 for the subpostmasters and subpostmistresses. Some of 21 the evidence in the first phase was truly distressing, 22 and painted a very vivid picture of the way that they 23 had suffered under this procurement. They have my 24 sympathy. 25 MR BLAKE: Sir, do you have any questions? 107 1 Questioned by SIR WYN WILLIAMS 2 SIR WYN WILLIAMS: Well, only to ask you, Sir Adrian: you 3 gave a brief summary of what might be called the hints 4 of recommendations in respect of PFI contracts in 5 general terms and what we might learn from the Horizon 6 PFI in particular. 7 If in due course I were to ask you to make a written 8 statement about such matters, would that appal you or 9 would you co-operate? 10 A. I would certainly co-operate, and it would be 11 a pleasure, sir. 12 SIR WYN WILLIAMS: Thank you. 13 A. I have some very firm views on what makes for a good 14 procurement and a bad procurement. 15 SIR WYN WILLIAMS: So I've gathered. 16 Thank you very much for making your witness 17 statement and for coming to answer the questions this 18 morning. I'm grateful to you. 19 A. Thank you. 20 MR BLAKE: Thank you very much, sir, we're back on Tuesday. 21 SIR WYN WILLIAMS: I believe that is the case, Mr Blake, and 22 10.00 on Tuesday, yes? 23 MR BLAKE: Thank you very much. 24 SIR WYN WILLIAMS: Thank you very much, everyone. 25 (1.17 pm) 108 1 (The hearing adjourned until 10.00 am 2 on Tuesday, 29 November 2022) 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 109 1 INDEX 2 PAGE 3 VINCENT GASKELL (sworn) ..............................1 4 5 Questioned by MS KENNEDY ......................1 6 7 Questioned by MR STEIN .......................38 8 9 SIR ADRIAN MONTAGUE (sworn) .........................50 10 11 Questioned by MR BLAKE .......................50 12 13 Questioned by SIR WYN WILLIAMS ..............108 14 15 16 17 18 19 20 21 22 23 24 25 110