11 October 2022
Hide video Show video
(10.30 am)
Housekeeping
Sir Wyn Williams: I thought we’d done away with standing, you’d all forgotten. As you can see, Mr Page is with me. Ms Eliasson-Norris is not, she can’t be here today but she’ll be watching the proceedings and obviously we’ll have a transcript and so will be just as au fait with what is going on as Mr Page and I.
I think we have a preliminary issue, if I can call it that, which I’ll ask Mr Beer to explain.
Mr Beer: Thank you very much, I was due to begin opening the issues that arise in Phase 2 to 7 of the Inquiry to you now and to speak for the next day and a half about them.
However, a pressing issue has arisen concerning Post Office Limited’s disclosure to the Inquiry, which I shall spend 15 minutes or so introducing now, and then give the floor to the legal representatives of those Core Participants who have indicated that they wish to speak about the issue.
As you know, sir, in order for it to be able to carry out its work, the Inquiry is reliant on those organisations who hold documents promptly to tell the Inquiry that they hold such documents and then to turn them over to the Inquiry, when requested to do so.
The Inquiry makes such requests by serving Rule 9 notices, that is notices pursuant to Rule 9 of the Inquiry Rules 2006, to produce such material to the Inquiry. The Inquiry may serve such notices seeking particular names, documents or broader classes of documents, perhaps those which concern an issue or which are within a certain date range or which are held at a certain location. When documents are handed over to the Inquiry there then begins a process, which is set out in the Inquiry’s protocols, to assess them for relevance to ensure that any information in them which ought not to be disclosed is redacted before disclosure to Core Participants, and then the documents are disclosed to the Core Participants on an electronic disclosure management system.
In order to be able to understand the document retention schemes and systems of relevant organisations, and the processes which they have undertaken in response to the Inquiry’s Rule 9 requests, the Inquiry may ask such an organisation to set these matters out in a comprehensive and complete manner in a witness statement, a so-called “disclosure statement”. This must be signed by a sufficiently senior and responsible individual within the organisation concerned. And it must be supported by a signed statement of truth. This allows the Inquiry to assess the steps that that organisation has taken in response to its Rule 9 request, in particular to see whether there are gaps or holes in the process.
The Inquiry has served 23 Rule 9 requests on Post Office Limited. It is right to say that the Post Office has disclosed to the Inquiry a vast quantity of disclosure, almost half of all documents held by the Inquiry.
On 8 May 2022, Scotland on Sunday published an article by Nick Read, Post Office’s chief executive officer. In that article, Mr Read stated that the Post Office had already provided over 3 million documents to the Inquiry. That statement was incorrect.
In a letter to the Post Office, dated 13 May 2022, the Inquiry noted that since its conversion back on 1 June 2021, the Inquiry had, in fact, received some 36,639 documents. Mr Read later apologised for this error in a letter to the Inquiry, dated 27 May 2022, and that letter of apology was published on the Inquiry’s website.
However, in the light of the discrepancy, and because of a number of tranches of disclosure had by then been completed by the Post Office, the Inquiry sought an interim disclosure statement from the Post Office.
On 27 May 2022, the Post Office provided a disclosure statement from Ben Foat, its general counsel. That statement which you’ve got, sir, as have the Core Participants, discusses hard copy documents at paragraphs 29 to 33 – just for your note, there’s no need to turn it up now, that’s tab 4, pages 16 to 17 – at a relatively high level.
It did not reveal the range of locations that we now know at which hard copy documents are, in fact, located nor the scale of the disclosure exercise connected with hard copy documents that the Post Office needed to undertake. Those matters have been revealed more recently.
To explain further, the Post Office has disclosed to the Inquiry that it is undertaking a significant organisation and review of a large number of boxes of hard copy documents and files held at various Post Office locations throughout the United Kingdom. Over a series of letters, the Inquiry tried to better understand the scope of the exercise, so that we could consider its potential impact on our work and our timetables. We also tried to understand the overlap between recent correspondence and the previous interim disclosure statement.
In short, the Post Office’s ongoing hard copy review and related correspondence puts the Inquiry in a very difficult position. We have limited information by which we can determine how much of the hard copy review is left to complete. We cannot know when we will be satisfied that the Post Office has completed its disclosure in response to all extant Rule 9 requests.
By way of example, some of Post Office’s recent hard copy document disclosures relate to Rule 9 requests that the Inquiry issued as far as back as October of last year. This makes it very difficult for the Inquiry to know when the Post Office’s disclosure will end and, therefore, very difficult for the Inquiry to work effectively and efficiently through the phases. We may be provided with highly relevant Post Office documents relevant to earlier phases after the hearings for those phases are complete. Those documents may be relevant to witnesses who have already given evidence and who may therefore need to be recalled.
However, the Inquiry has taken a number of steps as a result of recent events. Firstly, the Inquiry has requested a second interim disclosure statement from the Post Office, addressing their hard copy document review. In relation to the hard copy documents, this must address the following things:
A chronology of the work that the Post Office has undertaken to locate, upload, review and then disclose its hard copy documents since 1 June 2021, the setting-up date of this statutory Inquiry.
The work that remains as at the date of the interim statement.
It must set out all of Post Office’s hard copy repositories that the Post Office has been asked to expressly state the name of each such location.
The number of documents or, if the total number is unknown, to state as such and provide the number of boxes or files held at each repository.
An explanation of the upload review or other disclosure-related tasks being undertaken at each repository, in other words an explanation of the review progress made at each repository site.
That statement is due by no later than 4.00 pm on 18 October 2022, after the oral opening statements have been completed. At this stage, the Inquiry is minded to update representatives for Core Participants who have signed undertakings on or shortly after receipt of this second statement. The second step that the Inquiry has taken, as you will have seen from the correspondence before you, is that, as soon as the Inquiry was alerted to the issue, it requested that the Post Office should review the hard copy documents or material that was relevant to Phase 2 of the Inquiry, and prioritise the disclosure of that material.
For its part, the Inquiry has sought to prioritise such material for onward disclosure to the Core Participants when it is received from the Post Office.
It should be noted, however, that the Post Office has a large number of Phase 2 documents, with providers of documents at the moment and we have asked for some ambitious turnaround times in order to give disclosure ahead of relevant witnesses giving evidence.
Third, the Inquiry took the decision that it was necessary, in the interests of openness and transparency, to disclose all of the material about which I’ve just spoken – the disclosure assurance statement, the correspondence that occurred since then – to all of the Core Participants. That happened on Thursday last week and, of course, it was decided that it was necessary to allow Core Participants the opportunity to make any submissions on these important issues today and so, accordingly, we have allowed time for that to happen this morning.
Also the Inquiry assessed whether, in the light of the information now provided by the Post Office, some of which is not entirely clear, it was necessary to adjourn the starting date of Phase 2 of the Inquiry, to allow time for all of the disclosure of relevant material by the Post Office to occur.
The ongoing work by the Post Office, in relation to a large number of hard copy files and boxes across various locations, the scope and significance of which the Inquiry had not previously been made fully aware, means that it’s likely that some further documents that are relevant to Phase 2 of the Inquiry will only be disclosed during the course of the Inquiry. But what I would say is that the Inquiry has always made clear that the phases of the Inquiry are not hermetically sealed and witnesses can be called and re-called at later stages. Further witness statements can be sought and evidence can be read into the record at any time.
This Inquiry will not conclude until all Core Participants have met their disclosure obligations in full.
In our view, there is not sufficient reason to interrupt the good progress that this Inquiry is making. It may be that, over the lifetime of the Inquiry, Core Participants are provided with documents or with other information shortly before the hearings but this will be necessary to achieve or to help to achieve one of the Inquiry’s core aims: that is to establish what went wrong as swiftly as possible so as to hold those responsible accountable, to learn lessons and prevent this tragedy from occurring again.
Fifth, sir, we suggest, as your legal team, that you devote some time within Phase 2 of the Inquiry to investigating the adequacy of the disclosure that the Post Office has given to the Inquiry. It is, of course, something of a paradox, if not a matter which is dripping in irony, that I should begin this Inquiry by a discussion over the adequacy of the Post Office’s disclosure to the Inquiry when one of the very matters that the Inquiry is investigating is the Post Office’s past non-disclosure to the criminal and civil courts of this country.
We appreciate that issues such as this may reinforce the beliefs of many as to the sincerity and accuracy of statements made by the Post Office, as to its intention fully to cooperate with the Inquiry. However, we suggest that this is not a basis on which to lose the momentum which this Inquiry has gained, in particular because it is likely to delay and frustrate the important progress that is being made properly and fairly to compensate all of the victims of this scandal.
Sir, that’s all I say by way of introduction at the moment. You know, sir, who has suggested that they wish to make submissions now. I’ll now give the floor to them, beginning, I think, with Mr Stein.
Sir Wyn Williams: I think, if it hasn’t been made clear, I will hear the submissions in the order that you were each going to make your opening statements, if that makes sense. So I think that’s Mr Stein, then Mr Moloney, then Mr Henry, and I believe Ms Gallafent wants to speak, and that’s about it at the moment, though I will obviously ask any other Core Participants whether they wish to say anything.
It should also be clear to you all that I have expressed a provisional view about the suggestion that I should adjourn the Inquiry and I’m not inclined to do it. But that’s not to say that some flight of advocacy may not cause me to think hard and change my mind. But that is my provisional view, essentially for the reasons which Mr Beer has outlined in summary. So I’d be grateful if those who are going to speak this morning focus their submissions on why there should be a postponement. They will have a further opportunity, if I do not postpone, to address issues of disclosure when they open their cases formally.
So I would like to focus this morning on whether or not there should be an adjournment, all right?
So with those introductory remarks, Mr Stein.
Submission by Mr Stein
Mr Stein: Good morning, alongside Mr Jacobs, I am instructed by Mr Enright and his team at Howe+Co Solicitors and together we act for 153 Core Participants for this Inquiry.
Sir, we’re very grateful to you for allowing us the opportunity to address you on these disclosure issues, dripping with irony as they are. Our primary submission is that to go ahead now is to allow the Post Office yet again to dictate the pace and content of disclosure. That’s what they’ve been doing for 20 years and that’s what we suggest they’re trying now and continue to do with this Inquiry.
Our clients believe they have earned the right to have a full investigation into this scandal, and we strongly suspect that you agree. We suggest that, for a full investigation to take place, that instead of allowing the Post Office to disclose what it wants when it wants, that you should adjourn this Inquiry, with all of the inconvenience that that will cause, to everybody that is connected with this matter, until at least the disclosure process for these first few modules are complete.
The Inquiry has been more than fair, extending the time for the Post Office to provide hard copy documents. Sir, the correspondence that we first saw last week on Thursday – I think it was – on 6 October at 6.30 in the evening, tells us that the Inquiry extended the time for the Post Office to provide hard copy documents until 29 July 2022, and that’s an email dated 25 July 2022.
The response to that from Herbert Smith Freehills is a reminder to the Inquiry in its letter of 5 August that much of the material that the Inquiry has requested in recent Rule 9 requests is historical, making the point, say Herbert Smith Freehills, that in many instances the material is over 20 years old. Whilst that is absolutely true, perhaps we need to briefly remind ourselves that it is the Post Office that has been fighting this matter tooth and nail for very nearly a decade. It was in July 2013 that Mr Simon Clarke, a solicitor from Cartwright King, who had prosecuted many of the cases, explained to the Post Office itself that the Fujitsu witness, Dr Jenkins, the mainstay of many of its prosecutions, was not to be trusted.
It might have been thought that that put the Post Office on some notice, that they may need, at that stage, to gather the material that will be necessary to consider what has happened.
Sir, you’re also being told that at Chesterfield and other sites, in the letter from Herbert Smith Freehills dated 5 August, that the Post Office believes it can decide to what extent disclosure can or will take place. So I quote from the letter of Herbert Smith Freehills dated 5 August of this year:
“Notwithstanding that there is a risk that Chesterfield and other sites hold further material that is relevant to the Inquiry’s terms of reference, POL considers that to complete a comprehensive hard copy review of all of the material would neither be necessary or proportionate, bearing in mind the volume of material that would need time reviewed on site and the additional time and expenses that would be incurred.”
Understandably on behalf of the Inquiry, Ms Smith asks in her letter, dated 26 August 2022:
“On what basis have you determined that further hard copy reviews are not necessary in addition to not being proportionate?” making a clear point that it is not for them to make such decisions.
We’ve mentioned the fact that the Post Office perhaps should have been on notice considering the background of the matter, regarding the Clarke advices, but not only that: the Second Sight reviews; the judgment number 3 by Mr Justice Fraser in March 2019; judgment number 6 in December 2019; this Inquiry was established in non-statutory form on 29 September 2020; in April 2021, the Court of Appeal overturned convictions; and this Inquiry was then converted into the statutory Inquiry on 1 June 2021.
There could be no doubt at all that the Post Office has been on notice throughout all of that time that they need to gather all material that would be relevant to the purposes of all of those proceedings, and make sure that they were ready to deal with whatever investigation was required.
In the judgment of Lord Justice Holroyde in the Court of Appeal, the failures of investigation and disclosure were, in our judgment, so egregious as to make the prosecution of any of the Horizon cases an affront to the conscience of the court, repeating, really, the point that we have both made – myself and my learned friend, Mr Beer – that disclosure really has been at the heart of all of these issues.
Now, sir, the terms of reference and scope to this Inquiry state the Government wants to be fully assured that through the Inquiry there is a public summary of the failings associated with Post Office Limited Horizon’s IT system and says it will consider all other relevant evidence, listen to those that have been affected, understand what went wrong and assess whether lessons have been learned and whether concrete changes have taken place or are under way at the Post Office.
If you were sitting, sir, as a High Court judge in this situation, dealing with matters in litigation, our respectful submission is that there would be no hesitation. This would be a matter dealt with by way of an adjournment, an explanation that has already been asked for and a query as to how the costs should be borne.
Well, in principle, we submit that the same should apply here.
Sir Wyn Williams: Well, that’s what I find difficult, because I have all the flexibility of an Inquiry, not a sort of four-week time slot or whatever, which will mess up every other list, and so forth and so on. So I know that analogy is being made but I do think that, in an Inquiry, I have such a great deal of flexibility, compared with a normal court case, that we ought to forget about core procedures, Mr Stein.
Mr Stein: The analogy holds good, we say, because, in order to proceed at this stage, without the relevant documents having been disclosed, on which we can make judgments as to how to put forward questions, how to deal with openings, how to, in fact, consider the shape in terms of submissions to the Inquiry, the shape of the Inquiry itself, it goes to the questions of what has happened so far, from Mr Cipione any the IT expert that has been brought in, what material should go to him and his team that has considered it over, I think, many months in the summer.
So, sir, the difficulty is that the Inquiry may be, in fact, saying “Well, we can deal with it as we go along, this is a rolling Inquiry, we can catch up, we can go back to matters, re-call witnesses”.
Essentially, our clients say that is jumping to the tune that is being played by the Post Office.
Now, you’ve asked for flights of advocacy to consider the possibility of changing what is, so far, your mind. Well, I’m not sure I’m up to that. What I can, though, is do this, I can say this: the Post Office has denied subpostmasters and mistresses any choice in their lives. They demanded of them investment; they demanded of them application in their work; they demanded of them that they comply with a contract that effectively said that they must pay up, no matter whose fault it is.
They then, of those individuals, prosecuted them before the civil courts and the criminal courts and many went to prison. Their lives have been dogged by lack of choice from the Post Office. It is a significant matter that we ask you to take into account, sir, that their choice is to adjourn.
That doesn’t mean that this matter should necessarily be left today and we abandoned all of this equipment that’s been so kindly and time consumingly set up. What it does mean is that you could go ahead with what we can so far, by way of openings but then adjourn before we get to evidence.
So before I sit down, let me just read you three examples of the many emails that we have had concerning this particular question from our clients.
From Malcolm Simpson:
“I’m emailing you by return to fully support your actions following this development on the eve of the Inquiry. As you have stated in your letter to Sir Wyn, none of us is surprised by the Post Office’s actions. Just angry and disappointed that they have been playing the same old games with absolutely no respect for anyone, just protect the brand, despite their reputation being continually shredded because of their consistently behaving in the worst possible way for many years. Abject arrogance. Please keep fighting tooth and nail to bring these people to task.”
Rita Threlfall:
“I fully support and trust any action you feel necessary regarding the Inquiry. I’m mortified to hear of Post Office behaviour. Have we not been proven innocent? It is a living nightmare grieving for our past lives and some redress. Will we ever be treated fairly?”
John Stephens:
“The Post Office and their legal team’s behaviour comes as no surprise. The events of the last 20 years have clearly shown they have no intention to do anything but hide their actions from bottom to top up behind smoke and mirrors. Justice Fraser could not have been clearer on his opinion of their actions and we are yet again dealing with their duplicitous behaviour.”
Our clients want this Inquiry to proceed. They want this Inquiry to proceed with all due celerity. They want this matter to find who is responsible, who is accountable. They want it done properly. They want it done with this Inquiry in charge of disclosure, rather than the Post Office in charge of disclosure.
This is the only opportunity to have this particular matter looked at in its entirety. Our client’s point is: let’s start it right.
Sir, those are our submissions.
Sir Wyn Williams: Thank you, Mr Stein for powerful and succinct submissions.
Mr Moloney.
Submission by Mr Moloney
Mr Moloney: Thank you, sir.
Sir, this joint application is pursued with considerable regret. Until Thursday evening, our understanding was that the Inquiry was satisfied with the response to each of the Rule 9 requests and we don’t repeat the sorry history of the Post Office disclosure failings over the years. They are extensive, as outlined by my colleague Mr Stein, King’s Counsel. In that context, sir, the fact that disclosure issues have arisen again is surprising, only in the sense that it would have been hoped that Post Office would have got its act together by now, but no.
To illustrate that, sir, might we just go to one document that the Rule 9 process has generated, and I think this can go on the screen, yes, I’m grateful, to Mr Beer, King’s Counsel. It’s the “Minutes of the Audit, Risk and Compliance Committee of Post Office Limited” on 22 September 2020, and the reference for those who seek to find it later is POL00021462. These are the “Minutes of the Audit, Risk and Compliance Committee from 22 September 2020.
If we could please go forward to page 5 and to the bottom of that page, we see there at 4.8, this is the commencement of this topic within the minutes, which is:
“GLO/Freedom of Information (FOI) Requests/GDPR and data protection: The team is extremely busy dealing with requests related to Historic Shortfall Scheme and related/linked FOI requests. He advised of 39 FOI enquiries considered vexatious which would be reviewed accordingly.”
But over the page, more importantly, from September 2020:
“Of serious concern and Committee discussion, was the discovery of 31,000 boxes previously unknown to the wider organisation, which are being reviewed …
“The Committee questioned whether management had a handle on data management controls such as archiving, and remarked on the lack of accountability within POL.
“NR [that’s Nick Read, the group chief executive officer] remarked that this was an unacceptable incident and that he did not expect this to brought to the Committee’s attention by compliance. A paper on data controls is expected at GE for discussion.
“The Committee recommended a data amnesty be organised for all Post Offices be considered.”
Then if we just move forward to the action points on page 13, we see at 4.8 that the recommendation is that:
“A paper on data controls … is expected at GE for discussion”, following the identification of 31,000 boxes of data.
Over the page, 4.8 continued at page 14, that:
“The Committee recommended a data amnesty be organised for all Post Offices be considered.”
Sir, that was in the context and that was at a time when disclosure was at its height in respect of the criminal appeals. It provides context, we say, sir, for the Post Office’s latest citation of its administrative arrangements in response to the Inquiry’s requests.
We are years on from that unacceptable incident, just over two years on, from that unacceptable incident, and the Post Office has long known that this Inquiry would need its full and careful cooperation.
We say that an adjournment is necessary in the interests of preserving the integrity and effectiveness of this Inquiry.
Sir, to focus now on the main question you ask: why shouldn’t you stick to your provisional decision, rather than simply complaining about what Post Office has done and in the past and is doing now? To proceed to opening but, much more importantly, oral evidence in these circumstances creates a real risk of prejudice, both to Core Participants and to witnesses as well, sir, because material emerges which might undermine the points that I might, for example, wish to develop in questioning, material may emerge which would alter the whole structure of questioning by Core Participants, and coming back to witnesses is never ideal.
Indeed, paradoxically, sir, in an attempt – and I don’t say this is your motivation for your provisional decision, sir – but in an attempt to save time now, it may extend the time that you have to sit to hear evidence.
Moreover, perhaps, in keeping with the emails that Mr Stein has just read out, far from allaying rumour and suspicion, we say, with deep regret, that proceeding to evidence would, rightly or wrongly, potentially generate considerable speculation.
So in the context of the long period of time preparing for this Inquiry, and the period of time that this Inquiry will receive evidence, a short adjournment would, we say be proportionate. Our client’s, in the impact hearings, have told you, sir, how important this Inquiry is to them and they’ve put their faith in the Inquiry and your team and, sir, they’re impatient to know why the Post Office acted as it did. It is because the Inquiry is so important to them and to the public interest that our clients would ask that it be done right, as they might see it, and be seen to be done right, and that an adjournment at this time of a short duration would be proportionate.
That’s all we wish to say, sir.
Sir Wyn Williams: Thank you very much, Mr Moloney.
Mr Henry?
Submission by Mr Henry
Mr Henry: Thank you, sir.
The effect of this announcement has had a psychological shock upon the Core Participants that I, together with Ms Page and Mr Schwarz of Hodge Jones & Allen, represent and have had the honour to represent, because it, as it were, echoes appalling instances of injustice that occurred in their own lives.
Take the case, for example, of Ms Felstead, who was told that she could be part of the mediation process and then was told that she could not, and it transported her back almost 20 years to when, as a teenager, she was sent to a Young Offenders’ Institution. I suppose the reason why I stand on my feet now and I support my learned friends is for this reason: it’s to treat these people at last with dignity, a dignitarian approach, not a pragmatic approach.
I do not, sir, submit that you are motivated by saving time. You will take as much time as you need. What you’re concerned about is not losing momentum, but what is momentum if we set the wrong tone at the outset, if the angle of trajectory is distorted by non-disclosure? Then you may not reach your ultimate objective, the target that you, I submit, long for as well, because it is evident from everything you have said and done that you want this Inquiry to be effective and that you have been, notwithstanding your many years in the practice of the law and dispensing justice, shocked by what you have already heard.
I’m afraid I can’t use the analogy of a puppy because, of course, the Post Office is a far more sinister and, I regret to submit, malevolent animal, but you have to train an unruly participant and the only way you can train it is by actually not acceding to its continuing contempt for the process. It has demonstrated time and time again that it cannot be relied upon or trusted to comply with court orders. Once more, it has demonstrated that it is unwilling or unable to meet deadlines which have been set and it is utterly incorrigible.
So really, I suppose, my submission is Pavlovian, that yes, we do lose, to some extent momentum, but we will regroup and then we will attack this subject with even greater momentum and confidence, knowing that a marker has not simply been laid down but imposed upon an institution which seems incapable of acting fairly towards those it has maimed and marred.
As you know, sir, in the case of the Core Participants we represent, the Post Office exercised almost total power over them. It treated them contemptuously, it subverted the rule of law to suit its own agenda. It twisted, distorted and overrode vital processes in both civil and criminal courts, depriving the Core Participants of vital rights, which meant that the law could not protect them, which meant that the system of justice that we fondly believed would intervene, in fact, was enlisted in support of the Post Office’s vendetta, because it did flatten people who sought to challenge the narrative.
It is therefore now time, we respectfully submit, that control is imposed upon them, that they are treated with firm and obdurate – well, not obdurate, but with firm and rigorously imposed resolution that this must not be allowed to take place because, otherwise, I respectfully submit, as the history of the Horizon litigation in the High Court has shown and also the criminal litigation in the Court of Appeal, that the Post Office will fail to comply.
So out of dignity for those we represent, and for all the Core Participants represented by my learned friends, and also out of, I suppose, a Pavlovian imperative, you have to impose control now, sir, upon them. This is a seismic moment and I respectfully submit that your laudable desire for momentum has to be sacrificed to ensure that the Post Office do not continue to act in its arrogant and imperious way. Those are my submissions.
Sir Wyn Williams: Thank you, Mr Henry.
Is there any other Core Participant present who wishes to support the application for an adjournment, other than the Post Office? I don’t know what they’re going to do. Let’s see if there’s anyone else who wants to say anything for the moment.
Right. Well, then, I think Ms Gallafent, you wanted to make submissions.
Submission by Ms Gallafent
Ms Gallafent: Thank you, sir.
We’re grateful for the opportunity to respond to concerns raised by representatives of the postmaster Core Participants. The allegation made in the application that we received on Monday is that Post Office is continuing in what is alleged to be a pattern of late and partial disclosure of relevant documents as well as continuing to obfuscate attempts by others to attain those documents.
We emphatically deny that Post Office has deliberately sought to make late and partial disclosure of relevant documents. To put Post Office’s disclosure exercise in relation to hard copy documents into context, it’s important to note that, to date, first as Mr Beer indicated, there have been 23 separate Rule 9 requests issued to Post Office. In response, approximately 95,000 documents have been produced by Post Office to the Inquiry. That’s across more than 60 separate productions.
The production of those documents has taken place from a review pool of approximately 50 million documents and then a review of approximately 213,000 documents deemed to be relevant from the initial search.
The documents that have been disclosed were collated from an assessment of 173 Post Office databases, 201 Post Office SharePoint sites, eight external digital repositories and 248 physical locations.
It is understood that there is some suggestion by Mr Stein that Post Office’s disclosure has resulted already in delay to the start of this phase. For the avoidance of any doubt, our understanding, sir, from your public announcements in relation to the starting date of this phase, is it was because of the volume of disclosure that the Inquiry had received and would wish to consider, that of course is not only by Post Office but by other Core Participants, certainly not a failure to make disclosure.
What we say the correspondence between my instructing solicitors and the Inquiry team demonstrates is that Post Office did draw the Inquiry’s attention to some of the challenges faced in relation to hard copy documents in the interim disclosure statement dated 27 May 2022.
At the beginning of August, Post Office updated and proactively updated the Inquiry as to the progress that it was making in relation to hard copy documents. In mid-August, it indicated it would shortly be in a position to produce the first tranche of additional relevant hard copy documents, noting that fewer than 100 further responsive documents have been identified relating to the various Rule 9 requests that have been received at that time. There were 16 at that point.
By 6 September, the number of additional hard copy documents which were provisionally identified as relevant to Phase 2 was around 625. After review, documents which were identified as relevant to Phase 2, as well as Phase 4 were provided to the Inquiry on 13 September 2022.
The disclosure of those documents was made as soon as it reasonably could have been, having regard to the nature of the exercise being undertaken, and there was certainly no intention to delay their disclosure. There was equally no intention to obfuscate attempts by the Inquiry to obtain those documents. On the contrary, in correspondence, Post Office set out in detail the steps that it was taking in order to review potentially relevant documents so that it could make disclosure of any such documents to the Inquiry.
We understand from Mr Beer’s submission this morning that the Inquiry team felt that further detail could and should have been provided but that is something that was and can continue to be addressed both in correspondence and, of course, in the disclosure statement, the second interim disclosure statement, which will be supplied to the Inquiry on or before next Tuesday.
So we emphatically refute any suggestion that we have intended to obfuscate attempts by you, sir, and your team to obtain these documents.
There are two ongoing areas in which the exercise of identifying and reviewing potentially responsive documents has not yet been completed and these appear in the correspondence and let me summarise them.
The first concerns unindexed or inadequately indexed boxes and files. Now, this issue was first identified in the interim disclosure statement from May of this year, which identified the lack of indexing or the inadequacy of certain indexing, at the time understood to be about 5,000 documents at its offshore storage provider’s site. It’s said that those had recently been indexed as part of a program undertaken by Post Office, and the indices which had been produced were in the process of arranging to be reviewed.
It subsequently transpired that, in total, there were about 5,000 unindexed boxes and files but also about another 4,000 boxes where the indexing was inadequate, and Post Office via my instructing solicitors updated the Inquiry about that on 5 August of this year.
The new indices, including those where the indexing was inadequate, were provided to Post Office’s solicitors on a rolling weekly basis, in order for them to review each new index and identify if the boxes or files contained any documents responsive for the Rule 9 requests it had received by that date.
Having reviewed the new indices produced during August, Post Office’s solicitors identified an initial tranche of documents, which it considered might relate to Rule 9s, responsive to Phase 2 of the Inquiry and these, as I’ve indicated, were provided to the Inquiry in early September.
A tranche of 278-items was also identified in August which they considered may relate to Rule 9s responsive to Phase 3 of the Inquiry. Now, these documents are in the process of being scanned and, in some cases, reviewed and Post Office expects to be in a position to provide any relevant documents identified well in advance of the start of Phase 3 of the Inquiry.
In addition, after reviewing a further tranche of indices, received throughout the month of September – as I’ve indicated, they were provided on a rolling basis as the process continued – during the week commencing 26 September of this year, a further 111 boxes and 19 files of hard copy documents were identified from those new indices as being potentially relevant to Rule 9s that had previously been received.
Of those boxes and files, as at today’s date, 16 items, either boxes or files, have been identified as being potentially relevant to Phase 2, 89 potentially relevant to Phase 3 and 25 potentially relevant to Phase 4.
There have since been two supplemental indices received from the indexation exercises completed by Post Office. These have been reviewed for relevance to current Rule 9 requests. None of those documents have been relevant, identified as relevant to Phase 2.
For completeness, sir, I note that the indexing exercise being undertaken by Post Office is nearing completion. We understand there may be additional indices which it’s anticipated will be provided to those instructing me once finalised and they will be reviewed for potential relevance in the same way as previous indices.
Can I go back, then, to the additional 111 boxes and 19 files? It had initially been intended for these relatively recently identified items to be scanned before being reviewed. Post Office has now arranged for a team of five of its solicitors from Herbert Smith Freehills and a senior employee of Post Office to travel to the offsite storage provider’s site in Winchester tomorrow and Thursday, if necessary, to conduct the review in person.
The team will prioritise the review of any documents potentially responsive to Rule 9 requests in relation to Phase 2 and, if any are identified, Post Office will obviously provide copies of such documents to the Inquiry at the earliest possible opportunity.
The second area in which the disclosure exercise has not yet been completed, and again the background is set out in the correspondence, concerns files which were uploaded to the online document hosting platform Omnidocs. At the time of Herbert Smith Freehills’ letter of 6 September when these were referred to, more than 28,000 files had been uploaded, of which 17,862 had been identified, using a basic search, as containing the word “Horizon”.
It has subsequently proved harder than anticipated to review those files, due to the limited search functionality of the particular platform onto which they had been uploaded, and it was Post Office’s initial view, as communicated to the Inquiry on 29 September, that it would be unlikely to be possible to review the files in a timely or proportionate manner.
After further consideration, Post Office determined last week that it should review a sample of those documents to reach a more informed view of the extent to which these files might contain potentially responsive documents.
Its solicitors will therefore now be undertaking a review of 500 documents from the Omnidocs set, which they expect to be able to complete by next week.
On the basis of the outcome of that review, Post Office will be in a position to update the Inquiry as to whether it considers that the public interest in obtaining the information outweighs the cost and time, bearing in mind how important the information is likely to be. That, of course, reflects the test which is established by section 21.4(b) of the Inquiries Act, in relation to a claim by a person, it’s not reasonable in all the circumstances to require them to comply with a section 21 notice to produce a document in their custody or control. Of course, no such notices had been issued in relation to any of the Rule 9s to date but, in our submission, it is appropriate to have regard to the same threshold test when considering whether it is reasonable for Post Office to respond to a Rule 9.
Of course, we make no submissions at this time, prior to the sample, as to whether it would or wouldn’t be.
We fully recognise that if the sample exercise indicates that the public interest would outweigh cost and time, it would have to move with all possible expedition to complete an appropriate review of the full shade of files to minimise the impact on the oral hearings in Phase 2.
All of the issues that I’ve identified in terms of recent developments and updates from the end of the correspondence trail will, of course, be covered in the second interim disclosure statement due next Tuesday.
In these circumstances, our position is first that there is no basis for adjourning the hearings this week, and the application to do so should be declined.
Post Office would wish to reiterate it remains committed to responding fully to all Rule 9 requests, unless there are particular circumstances where it considers it’s not reasonable to do so, which it will then invite the Inquiry to consider and offer a view on. It will continue to flag up to the Inquiry in correspondence any difficulties or potential delays with disclosure exercises, not limited to Phase 2 but otherwise, and seek to find practical and workable ways through.
It very much apologises for the fact that it is necessary for this issue at the take up time at the outset of the hearing but we recognise, of course, the concerns and the rationale for them raised by my learned friends.
We fully recognise that the Inquiry will wish to review the situation once it is in receipt of the second interim disclosure statement and we note Mr Beer’s suggestion that some time during the forthcoming phase can be taken on the issue. Obviously, we don’t object to this action. In the circumstances, we respectfully agree that there is no basis on which to lose the momentum provided by the start of this phase, in particular not to adjourn today, and we say matters thereafter are matters that can properly and fairly be considered during the course of Phase 2 and, of course, the Inquiry will make such directions or such decisions as it considers appropriate in due course.
Thank you.
Sir Wyn Williams: Thank you.
Before I ask Mr Beer if he wishes to say anything, are there any Core Participants present who wish to support the view that I should not adjourn or postpone the hearings? Right.
Mr Beer.
Ruling
Sir Wyn Williams: Let me say at once that, in reaching my decision, I have taken account of the representations which I have read from a number of Core Participants and, as I’ve said, the full document begins with representations from Ms Palmer and ends with representations from Ms Hazzleton.
The representations support an application which has been made to me to adjourn or postpone the start of Phase 2 of the Inquiry. I say start because it became clear, during the course of the oral submissions, that at least some of those supporting a postponement would be content for opening statements to be made but do not accept that I should begin to hear evidence.
It is obviously regrettable that disclosure of documents potentially relevant to Phase 2 is not yet complete. That seems to me to be clear from the correspondence which everyone has seen and from the oral submissions which I have heard today.
That was not a surprise to me, in the sense that my team has been reviewing this issue over the last number of many weeks, and my legal team has been considering over that time, whether or not it was appropriate to start these proceedings on this date. As Mr Beer explained in his opening remarks, notwithstanding the fact that disclosure in relation to Phase 2 was not complete, he and the legal team had reached the conclusion that, nonetheless, it was appropriate and reasonable to begin today.
I issued a provisional view to the same effect after I had received the written representations made on behalf of the Core Participants who sought an adjournment.
I say now that, despite the powerful advocacy which had been displayed before me this morning, I am still of that provisional view.
A public inquiry is a flexible creature. I think those of you who have appeared before me regularly know that I am a flexible creature. I do not think it appropriate to be hamstrung by straitjackets or anything of the like. The issue for me, as I see it, is whether or not I can proceed fairly, reasonably and appropriately, notwithstanding that disclosure is not yet complete. I believe that with appropriate management and assistance from everyone, which includes my legal team and every other legal team, I can proceed.
The plain fact is that when one is investigating events which occurred 20 years ago, as we are in this phase, it is of no surprise that sometimes there are glitches in relation to disclosure. I appreciate that there are many people listening to my words who might wonder whether or not I am going to control disclosure in the way that I have been invited to do so.
Let me assure them that I do intend to control disclosure. No one should be under any illusions. The obligation to disclose relevant documents is ongoing and will persist throughout this Inquiry until the very day when I hand over my report to the Minister.
I wish to make it clear that, if I reach a conclusion at any time that there is either accidental or, even worse, deliberate, non-disclosure of relevant documents, I will use all the extensive powers at my disposal to obtain the documents which have not been disclosed.
Everyone should understand, and I know that the lawyers do understand, that proper disclosure is crucial to the success of this Inquiry.
Those are general remarks. I have considered whether we have reached a point where we need to pause because there is a danger that the non-disclosure which has occurred to date will interfere with my aims for this Inquiry. For the reasons essentially articulated by Mr Beer, I have reached the conclusion that that stage has not been reached.
I am satisfied that I can manage disclosure. If it is necessary, I will set aside days in this phase or any other phase to consider disclosure as a discrete topic.
If it is necessary, I will attribute blame to any party where that is justified in relation to non-disclosure. I will not deflect from being critical if it is necessary to do so. So this is a crucial part of this Inquiry and I wanted to lay down those guidelines, as I will call them, in the hope that I never need to get to directions or even more draconian use of powers.
I also have to bear in mind the effect that an adjournment or postponement would have at this stage of the Inquiry. With respect to those who suggested that it might be a short adjournment, I beg to differ. The reality of an adjournment would be a reasonably lengthy period of time while I satisfied myself that every single relevant document had been disclosed. That would have a very significant impact upon those timetables which have already been drawn up with considerable care by my team and it would have a very significant impact, therefore, going forward on every other phase of this Inquiry.
While expediency about timetables is not an overriding consideration, it is an important consideration for all those subpostmasters – and I use the term generically, as we always have – because they do need answers as soon as is reasonably practicable, as I have been at pains to stress in virtually every public statement I’ve ever made.
So balancing all the relevant considerations and despite the powerful advocacy of those promoting a postponement, I’ve decided we should carry on. But let me make one thing clear: the disclosure duty is a continuing one and I will supervise it with as much vigour as I can. Throughout my years as a lawyer and judge, I have been in too many cases where disclosure difficulties have started to become, let’s say, very difficult to reconcile with proper objectives. I will not allow that to occur in this Inquiry.
Thank you.
Mr Beer?
Opening statement by Mr Beer
Mr Beer: Sir, thank you. If I may, I’ll open the remaining phases of the Inquiry.
The long fought –
Sir Wyn Williams: Actually, Mr Beer, before you start and to save my aged hand, there used to be in this Inquiry almost a more or less contemporaneous transcript of what you are saying, which saved me from having to make extensive notes, even though in an old-fashioned way, I used to try. Is that facility still available?
Mr Beer: Sir, it is and it should be on the screen in front of you, I think. I’ll just pause.
Sir Wyn Williams: Even pressing the button, which is about as much technology as I’m good for, doesn’t seem to bring that up.
Mr Beer: I’ll pass no comment on that last remark, other than to recognise it. I wonder whether somebody from RTS could help.
Ah, I understand that you’ve got an evidence – sir, you’ve got a transcript screen not an evidence screen. That will be rectified.
Sir Wyn Williams: So at least before lunch I have to resort to the old-fashioned note taking, Mr Beer?
Mr Beer: Well, not completely, because obviously a transcript will be available in due course.
Sir Wyn Williams: Fine.
Mr Beer: As I was saying, the long-fought group litigation in Bates & Others v the Post Office came to an end in December 2019.
On 26 February 2020, the then Prime Minister, Boris Johnson, committed to getting “to the bottom of the matter” and by 10 June that year, the Government had confirmed that there would be what was first called an independent review to be led by an independent chair.
The Post Office Horizon IT Inquiry was first established in a non-statutory form on 29 September 2020. On 5 October 2020 the Inquiry published an open letter, aimed at seeking information from the public, and started discussions with individuals and organisations, and the process of gathering of evidence.
A number of developments occurred in the months that followed. Most significantly, on 23 April 2021, the Court of Appeal Criminal Division handed down its judgment in the case of Josephine Hamilton & Others v The Post Office, which quashed the convictions of 39 subpostmasters where the reliability of Horizon data had been essential to their prosecution.
As you did, sir, I should say at this point, I will refer to “subpostmasters” throughout this opening. That is to be taken to include subpostmistresses, managers and assistants.
That judgment, as I will set out in due course, included new information which had not previously been disclosed, including what has become known as “the Clarke advice”, namely advice to the Post Office from a barrister, Simon Clarke, which highlighted serious failings with expert evidence that had been relied upon by the Post Office in prosecutions and further failures in respect of the duty to record and to retain information. On 19 May 2021, it was announced that, having digested the judgment of the Court of Appeal and returned to the judgments of Mr Justice Fraser, you had requested the conversion of the Inquiry into a statutory Inquiry and had suggested amendments to the terms of reference.
It was considered that the powers available to a statutory inquiry, including the power to compel evidence and hear evidence on oath, were necessary to support a proper assessment of the relevant facts. The Inquiry was therefore formally converted from 1 June 2021. Since that date, the Inquiry has published a range of protocols, including a protocol for providing oral evidence at hearings, a protocol on witness statements, a protocol on redaction, anonymity and restriction orders, amongst others.
Two assessors have been appointed as you know, David Page of Solirius Consulting, who sits next to you as an assessor in relation to digital product management, IT strategy for large and complex technical projects, usability and associated business processes, in essence our technical assessor; and Erika Eliasson-Norris of Beyond Governance, as a governance assessor. As you have said, sir, she joins today remotely.
On 17 November 2021, a completed list of issues was published, following a process of consultation with Core Participants and other interested persons and in the light of written and oral representations. The next day, the Inquiry sought permission to investigate matters relating to the devolved administrations, following concerns that had been raised, during the consultation, on the list of issues, as to whether and to what extent there had existed divergences in the policies and practices adopted within the four countries of the United Kingdom when taking action against subpostmasters.
Permission was granted by the then Minister for Small Business, Consumers and Markets, Paul Scully MP, on 30 March 2022.
Meanwhile, we began the human impact hearings on 14 February this year. Those hearings focused on the human impact of the adoption, implementation, operation and continued use of the Horizon System, and the human impact of the approach taken by the Post Office to subpostmasters when the system showed shortfalls, including their suspension, the termination of their contracts, the closure of their branches, proceedings for the recovery of so-called shortfalls, bankruptcy proceedings, criminal investigations and, ultimately, criminal proceedings.
The reason that we commenced the hearings of the Inquiry by listening to the accounts of those who were and continue to be affected by the flaws in Horizon and the flaws in the Post Office’s approach to it is that they must be front and centre of this Inquiry.
As I set out in my opening to this phase, although the underlying subject matter of the Inquiry is information technology, this Inquiry is not and will not become a dry technical investigation into an IT project gone wrong.
That’s because it’s an Inquiry that’s actually about people: about people whose mental and physical health has been impacted; about people whose marriages and partnerships have deteriorated and failed; about people who thought about taking their own lives; and, in some cases, who took their own lives.
Between February and May of this year, we heard from those affected. The Inquiry travelled to Cardiff, to Leeds, to Glasgow and to Belfast and heard powerful accounts from victims and, in some cases, family members, including from those whose loved ones did not live long enough to see their convictions overturned.
In terms of disclosure, the Inquiry has sent 91 requests for information under Rule 9 of the Inquiry Rules 2006, requested witness statements from 309 witnesses and disclosed to Core Participants a total of 10,237 documents. This work will, of course, continue over the coming months.
Along with Julian Blake, Catriona Hodge, Sam Stevens and Ruth Kennedy, I act as Counsel to the Inquiry. As you may know but others may not, our role is to assist you in the conduct of these investigations. We are entirely independent and impartial and do not represent the interests of any of the Core Participants in the Inquiry or indeed of any other person.
Also appearing today – and I’ll only introduce them once in the course of the Inquiry – and in no particular order, are the following representatives of Core Participants:
Tim Moloney, King’s Counsel and Angela Patrick, instructed by Hudgell Solicitors for 64 subpostmasters.
Sam Stein, King’s Counsel and Chris Jacobs, instructed by Howe+Co for 153 subpostmasters.
Edward Henry, King’s Counsel and Flora Page instructed by Hodge Jones & Allen for five further subpostmasters.
Kate Gallafent, King’s Counsel, instructed by Herbert Smith Freehills for the Post Office Limited.
Richard Whittam, King’s Counsel and Tim James-Matthews instructed by Morrison Foerster for Fujitsu.
Paul Mertens instructed by Eversheds Sutherland for UK Government Investments or UKGI.
Nick Chapman, instructed by the Government Legal Department for the Department for Business, Energy and Industrial Strategy or BEIS.
Samantha Leek, King’s Counsel instructed by Mishcon de Reya for Paula Vennells.
Phase 2. I now turn to address the matters which will be examined during the Phase 2 hearings, which commence with evidence next week.
As you know, sir, Phase 2 of the Inquiry is concerned with the first iteration of Horizon, known as Legacy Horizon. During Phase 2, the Inquiry will examine how this first iteration of the Horizon IT system came into being and will be focused upon decision making in relation to the following four overriding themes: firstly, the procurement of the Horizon IT system; secondly, its design and development; thirdly, the Treasury review and cancellation of the benefit payment card or BPC; and, fourthly, the acceptance and rollout of Horizon.
The significant modifications which were later made to Horizon will be examined in further depth during Phase 3 of the Inquiry. You will also hear during Phase 3 from witnesses who were involved in the early decision making relating to Horizon and who continue to have direct involvement in the operation and management of the system after it was rolled out to Post Office branches nationwide.
Their evidence will, of course, be of relevance to the issues that you will be determining as part of Phase 2 as well.
Can I now then turn to address each of the four of the Phase 2 themes that I mentioned in turn.
Procurement of the Horizon IT system. The plan to bring modern technology to the entire Post Office Network was first conceived in the early 1990s. At that stage, the Post Office consisted of three principal businesses: Post Office Counters, Royal Mail and Parcelforce. Post Office Counters was the business responsible for running the national network of Post Offices which, at that time, comprised of approximately 19,700 branches.
Over 95 per cent of those branches were run as private businesses by agents of the Post Office, known as subpostmasters. For many years, Post Office Counters had worked in close partnerships with the Benefits Agency, facilitating the payment of state benefits over the counter at Post Office branches nationwide. In the early 1990s, the method of paying state benefits via the Post Office was paper based and reliant upon the use of order books or green giro cheques. It was expensive for the Benefits Agency to administer benefit payments in this way and the system was considered to be very vulnerable to fraud.
An alternative method of paying state benefits directly in relation to a recipient’s bank account had already been established by the Benefits Agency. This method was known as, with capital letters, Automated Credit Transfer, ACT, and was less expensive than paying payments via the Post Office. Transferring payments to recipients via ACT would have reduced the administrative cost borne by the sponsoring department of the Benefits Agency, which at that time was known as the Department of Social Security, the DSS.
However, the income associated with the payment of state benefits accounted for more than one-third of Post Office Counters’ business, and was considered at the time to be essential to the survival of the Post Office Network. For this and other reasons, the option of making payment via by ACT compulsory was not pursued by the DSS. Instead, Post Office Counters and the DSS embarked upon a joint venture to automate the payment of state benefits over the Post Office counter, via the use of an electronic payment card.
The Benefits Agency acted as the sponsor of the automation project on behalf of both the DSS and the Northern Irish Department of Health and Social Care, as well as other agencies. From the perspective of the Benefits Agency, the principal objectives of the project were: (i) to eliminate encashment fraud; (ii) to reduce their administrative costs substantially; and (iii) to improve their financial accounting and management of information.
Although one of the principal objectives of Post Office Counters was to retain the business of the Benefits Agency, the automation project had the broader objective of modernising the Post Office Network. At the time, Post Office Counters operated a paper-based accounting system across its network of, as I’ve said, nearly 20,000 branches, which was costly to reconcile and was considered to act as a barrier to attracting new business.
Automating the sale of Post Office products and Post Office Counters’ other client transactions was intended to increase efficiency, improve business competitiveness and enable Post Office Counters to attract new clients.
The process of procuring a new IT system to automate the payment of benefits at Post Office branches was managed by a joint programme team with significant input from external lawyers and consultants. The joint programme team was staffed with representatives drawn from the two project sponsors: Post Office Counters and the Benefits Agency. It was led by a programme director, Andrew Stott, who had previously worked as a senior official within the DSS.
Mr Stott reported to a programme steering committee, which comprised of the chief executive and finance director of the Benefits Agency, the managing director and commercial/resources director of Post Office Counters, as well as senior officials drawn from Her Majesty’s Treasury, as it was then known, the Department of Trade and Industry, the DSS, the Northern Irish office, the Private Finance Executive. It was the preference of the government of the day that funding for the project should be sought from the private sector and the new IT system therefore be procured under the private finance initiative.
The procurement process was launched on 30 August 1994 by the publication of a notice in the Official Journal of the European Communities. For the transcript, but not needing to be displayed, the reference to that notice is [WITN03770102]. One of the businesses which expressed an interest in the project was International Computers Limited, ICL, a British manufacturer of computer hardware and software, and a provider of computer services, which was partially owned by Fujitsu Limited, a Japanese information and communication technology company.
ICL established a subsidiary company known as Pathway, in order to bring together a consortium of businesses which could jointly bid for the project. The other members of the consortium included De La Rue and Girobank.
The process of shortlisting the potential service providers which had responded to the OJEC notice was undertaken in the autumn of 1994 and resulted in the selection of five companies: BT Syntegra, Cardlink, EDS, IBM and Pathway.
On 13 April 1995 the project sponsors issued to the five shortlisted companies a statement of service requirements known as an SSR, which, amongst other things, described the functional requirements which each of the sponsors expected the project to deliver. That SSR can be found but needn’t be displayed at POL00028135.
The SSR summarised the principal tasks of the service provider as follows: (i) to provide, operate, support and manage an IT infrastructure which would enable the automation of services at Post Office’s; (ii) to automate the end-to-end benefit payment system; (iii) to integrate with the computer systems of Post Office Counters and the Benefits Agency as well as those of their agents and clients; (iv) to migrate the project sponsors’ existing automated systems; and (v) to integrate Post Office Counters, the Benefits Agency and third-party software applications.
Each of the five shortlisted suppliers’ responses to the SSR were assessed by an evaluation team against three sets of criteria, addressing the characteristics, the viability and commercial aspects of the suppliers’ proposals.
That exercise can be found but needn’t be displayed at POL00028282. At the conclusion of their assessment, the evaluation team produced a report for the members of the Project Evaluation Board to assist them in selecting a final shortlist to move to the more detailed demonstrator and negotiation phase of the process. That can be found but needn’t be displayed at POL00028149.
The evaluation board comprised representatives of Post Office Counters, the Benefits Agency, the Social Security Agency of Northern Ireland. It was chaired by Robert Peaple, executive director of Post Office Counters.
In its report to the evaluation board, the evaluation team identified two principal areas of risk in Pathway’s proposed solution. The first area of risk was technical and associated with Pathway’s proposal to adopt a distributed solution which would enable transactions to be processed at the Post Office counter, without requiring continuous connection to a central processing system. The technical risks which were seen to be inherent in such a solution were twofold: firstly the greater complexity of the system software required to manage such a system; and secondly the greater complexity of administering benefit payments at any Post Office branch, other than the one specifically nominated by the recipient.
The second area of risk identified by the Evaluation Team was commercial and related to Pathway’s apparent refusal to accept any risks for financial losses arising from fraudulent copying or counterfeiting of the benefit payment card. However, the Evaluation Team did not consider that this commercial risk ought to preclude Pathway from proceeding to the demonstrator and negotiation phase of the procurement exercise, where the method of card holder verification could be examined more fully.
Ultimately, in July 1995, Pathway was shortlisted alongside Cardlink and IBM to continue to the next phase of the exercise.
The demonstrator and negotiation phase of the procurement exercise was intended to serve three principal objectives: firstly, to clarify and define more precisely the requirement of the two sponsors; secondly, to obtain a clearer understanding of the suppliers’ proposed solutions to them; and, thirdly, to prepare draft contracts defining the rights and obligations of the right contracting parties.
These activities were underpinned by a Service Provider Risk Register, which was regularly reviewed and updated as new risks in the suppliers’ proposed solutions were identified and steps taken to mitigate or to eliminate those risks.
That can be found but needn’t be displayed at DWP00000036.
The risks associated with each of the suppliers’ proposed solutions were ranked according to their severity, with the most critical risks being classified as category A. That can be seen but need not be displayed at Fujitsu, that’s FUJ00078056. Each of the shortlisted suppliers was notified of the risks identified in their solutions and all afforded an opportunity to respond to the concerns raised by the programme team, either in writing or during their weekly demonstrator meetings with representatives of Post Office Counters and the Benefits Agency. You can see some examples of those, we needn’t display them, at FUJ00078002 or WITN05970140.
You’ll be hearing evidence from Jeremy Folkes, a software developer employed by Post Office Counters who participated in this demonstrator work stream. He will say that Pathway appeared defensive in its meetings with the project sponsors and appeared keen to avoid scrutiny. We will see that in his witness statement at WITN05970100 at paragraph 43.
You will wish to consider, Chair, whether these observations mirror trends noted later in the development of the project.
In order to proceed to the final stages of the procurement process known as the Invitation to Tender, each of the service providers were required to revise and develop their solutions to ensure that they met or exceeded: firstly, the minimum service requirements acceptable to the project sponsors; secondly, the minimum requirement for partnership with Post Office Counters and the development of new business opportunities; thirdly, sufficient transfer of risk for an acceptable risk profile, in particular relating to fraud; and, lastly, an acceptable funding method and finance structure. One can see those criteria at POL00028297.
One of the principal methods by which the joint programme team determined whether or not these hurdles had been met was by reference to the Service Provider Risk Register, which I’ve mentioned already, the stated intention of the project sponsors was to exclude from the invitation to tender any shortlisted suppliers who had category A risks or “an unacceptable profile of other risks” outstanding at the conclusion of the demonstrator and negotiation phases of the procurement. We can see that that is so, ie, as a stated intention to exclude suppliers who had any category A risks from POL00028297.
An assessment of Pathway’s proposed solution in November 1995 identified four outstanding category A risks, as well as 20 category B risks that were described as “significant”.
The greatest area of concern to the joint programme team related to Pathway’s proposed reliance on a software platform known as Riposte, which had been developed by a small software company called Escher based in the United States. The project sponsors had serious concerns about the scalability of Riposte and perceived that ICL did not have a sufficiently deep understanding of the software to answer their questions. It was noted that Pathway had the highest number of category A and B risks of the three shortlisted suppliers and that further work needed to be done to clear the risks if Pathway was to be permitted to proceed to the invitation to tender. One can see that at POL00028148.
The demonstrator and negotiation phase came to an end in February 1996. An analysis of the Service Provider Risk Register carried out on about 19 February 1996 suggested that Cardlink and IBM had no outstanding category A risks and accordingly should not be included from the invitation to tender.
By contrast, Pathway had two outstanding category A risks and was therefore considered not to qualify for the final stage of the procurement process.
I think we could look at that document usefully, please, at POL00028150. You’ll see the date in the top right-hand corner, 19 February 1996, and if we go to page 3 of the document, one can see a table here. This is of the Pathway risks. You’ll see that under the column “Risk severity A” there were two outstanding. I should say this is not our highlighting. The highlighting is as it appeared in the document that we have been given. And you will read, at paragraph 2.3.2:
“Pathway has two A risks. In accordance with the procurement policy concerning A risks, the Programme should not invite Pathway to tender.”
You’ll see a scribbled note on the right-hand side, “now changed and cleared 26” – and that’s a “2”, I think – “told at evaluation board.” We will come to that in a moment.
So the two outstanding category A risks recorded against Pathway related, on the one hand, to Pathway’s continuing refusal to accept any risk for financial losses arising from fraudulent copying or counterfeiting of the benefit payment card, and on the other, to Pathway’s proposed financial and funding arrangements which were considered to represent an unacceptable risk to the project sponsors.
We can take that document down now.
In addition to the two outstanding category A risks, Pathway had nine outstanding category B risks reflecting a higher aggregate risk profile than either of its other two competitors. The significant category B risks associated with Pathway’s proposed solution related to a variety of technical and contractual issues arising from Pathway’s reliance on the Riposte software produced by Escher, and the adequacy of its proposals to ensure payment security and to eliminate fraud. As I’ve said, in the light of the Pathway’s outstanding category A risks, it was recommended that the consortium be not invited to tender.
The ultimate decision as to whether any of the shortlisted suppliers ought to be excluded from the invitation to tender rested with the Evaluation Board. Prior to the meeting of the board on 26 February 1996 – remember, that was the date in that handwritten comment – further efforts were made by Pathway to mitigate its two outstanding category A risks, and contractual negotiations with all shortlisted suppliers continued to take place. We can see that, but it need not be displayed, at POL00028290.
Although Pathway was deemed by the time of the meeting on 26 February to have cleared the hurdle of satisfying the minimum service requirements acceptable to the project sponsored, there remained concerns on the part of the joint programme team about Pathway’s management and its technical capability which raised fears about their ability to deliver the contracted services on time and to quality. We can see a reference to that, a record of that, in POL00028291.
At a meeting of the Evaluation Board on 26 February 1996, which was chaired by Robert Peaple, the board expressed “severe reservations about the number and significant risks against the Pathway proposal” and “questioned whether the cumulative effect would not lead to potentially late delivery and/or operational problems, and as such, was cause for Pathway to be excluded from the invitation to tender.”
We can see that, and I’d ask for it to be displayed, at POL00028288.
This is a note of the meeting on 26 February of the Evaluation Board. You can see that in the fifth line in the heading the Chair of it is Robert Peaple, Bob Peaple, and if we go forward, please, to page 3, and look at paragraphs 2.7 and 2.8:
“With respect to the certificate for Pathway, the Evaluation Board expressed severe reservations about the number of significant risks against the Pathway proposal. It questioned whether the cumulative effect would not lead to potentially late delivery and/or operational problems, and as such was cause for Pathway to be excluded from the invitation to tender. Particular concerns were expressed about the card technology with shortcomings in the associated management of fraud and about the dependence on Escher as a small company subcontracted to Pathway.”
Then at 2.8:
“Tony Johnson explained that these issues had indeed been given much consideration by the demonstrator team. It had concluded that the Pathway solution was not fundamentally flawed and that from a strictly technical viewpoint the hurdle had been cleared.
“In its overall decision, the Evaluation Board might wish to take into account the cumulative effect of any contractual issues with those from the requirements area.”
Having been informed by the programme team that Pathway’s two outstanding category risks were capable of being cleared following further contractual negotiations, the Evaluation Board concluded that the other outstanding risks registered against Pathway’s proposed solution did not warrant their exclusion from the Invitation to Tender and therefore, on 29 February 1996, the Invitation to Tender was issued to all three of the shortlisted suppliers.
The final stage of the procurement process was known (again, capitals) as the Evaluation and Selection Phase. Its purpose was to make a final assessment of the relative merits of each of the shortlisted supplier’s bids and ultimately to lead to the award of a contract, provided that a suitable provider could be identified.
The evaluation process comprised of three principal work streams: first, a review of the extent to which the suppliers’ bids complied with the requirements of the Invitation to Tender (this was known as contract assurance); secondly, an evaluation of the financial impact of awarding the contract to each of the three shortlisted suppliers known as “the financial evaluation”; and lastly, thirdly, an assessment of the characteristics and viability of the suppliers’ proposed solutions, known as the “Value Factor Assessment”.
In addition to these three aspects of the evaluation process, assessments were made both of the suppliers’ partnership capabilities and also of the extent to which their bids made provision for risk transfer that was sufficient to comply with the requirements of the private finance initiative.
Paul Rich, who was then the partnership development director of Post Office Counters and a member of the Evaluation Board, will be giving evidence to the Inquiry in which he will say that the level of risk transfer mandated by the private finance initiative was “a critically important criterion” as neither of the project sponsors had the resources to fund the design, manufacture, installation and operation of the new system, and wished to transfer those risks, together with the risk of fraudulent encashments, over to the supplier. We will see that at WITN04030100 at paragraph 33.
The Value Factor Assessment, capital letters, represented the key area of qualitative analysis and was initially undertaken after the Invitation to Tender, but prior to receipt of the suppliers’ bids. It drew heavily on the knowledge and understanding of the suppliers’ proposed technical solutions which the project team had gleaned during the demonstration and negotiation phase.
We can see a table at POL00028294. If we go forward, please, to the third page of that document, in tabular form, we can see that out of the three shortlisted suppliers, Cardlink, which occupies the second and third columns, obtained the highest overall score, and that although Pathway’s performance, which are the last two columns, was broadly aligned with its competitors in devising a solution which was acceptable to the project sponsors and their agents, it scored substantially less against a number of important matrices including credibility of delivery, management, capability, reliability and support, and stability and coherence.
So overall, in qualitative terms, Pathway was assessed by the programme to be the weakest of the three candidates.
That can be taken down, please.
On 21 March 1996, each of the three shortlisted suppliers submitted their responses to the Invitation to Tender. Upon analysis of their bids, it transpired that each had been priced at a level which was higher than the indicative figures which had been produced at the time of their original proposal and risked rendering the sponsors’ respective business cases unviable. In the light of this risk, an invitation to re-tender was issued to all three candidates on 16 April 1996, and their revised bids were provided to the programme on 22 April 1996.
The process of financial evaluation was carried out by the Joint Project Team upon receipt of the suppliers’ revised bids. Its purpose was to establish the quantitative differences between the bids in purely financial terms. Cardlink emerged from the evaluation as substantially the most expensive of the three suppliers, with IBM marginally outperforming Pathway on the total costs of its offer. We will see that Pathway came in as the cheapest at POL00028153.
In tandem with this financial evaluation, a review of each suppliers’ re-tender was conducted by something called “contract assurance”. Of each of the three suppliers, Pathway was found to have assumed the greatest level of risk for fraud associated with card counterfeiting, hacking, and customer validation. According to the contract assurance team, the limits which the other two candidates had placed on the scope of their contractual liability for fraud rendered their bid “difficult, if not impossible, to accept”. It was the conclusion of the contract assurance team that Cardlink should not be awarded the contract “at any price”. Of the remaining candidates, it concluded that Pathway ought to be preferred to IBM, unless IBM’s bid offered “a considerable price advantage”.
We will see that at POL00028152.
The process of selecting the successful candidate commenced on 26 April 1996 at a meeting attended by six representatives from the Evaluation Team. You will be hearing from David Miller, a deputy director of the programme in 1996, and a member of the Evaluation Team who attended that meeting. Mr Miller states he was unaware at any time of any issues with ICL’s ability to assist with the automation process. That’s in his statement at WITN03470100 at paragraph 24.
You, Chair, will wish to examine what might be a surprising assertion in the light of the strong concerns that have been earlier articulated by the programme about Pathway’s overall management capability, its failure to demonstrate leadership within the consortium, and its ability to deliver services to time and to quality.
At the meeting on 26 April 1996, the Evaluation Team concluded that Cardlink ought to be eliminated from the competition. Of the two remaining candidates, IBM’s performance in the qualitative Value Factor Assessment remained substantially better than Pathway’s, despite Pathway’s scores improving as a result of the re-tender. Moreover, these two suppliers were virtually equal on price. However, IBM’s contractual stance, in particular on fraud risk transfer, was deemed to place it at a material disadvantage to Pathway and its bid was therefore regarded as non-compliant with the private finance initiative.
If we can turn up POL00031153, and turn to page 22, please.
Back a page, please. Yes. If we look, please, at 9.7, the record of the approach and conclusions of the Evaluation Team was that the group recognised that an award to Pathway “would imply a need for a proactive management stance” – that’s a phrase that we will see quite a lot in Phase 2 of the Inquiry – “would imply a need for a proactive management stance by sponsors, notwithstanding the improvement noted by the Contracts Stream ^ is this Contracts Team ^ since the restructuring immediately prior to the Invitation to Tender issue. It would also require sponsor staff to work closely with Pathway on fraud prevention measures, although given the changes on fraud risk made by the other two bidders in their re-tenders, most of this work was likely to be required whichever supplier was chosen.”
And therefore we can see at 9.9:
“The group therefore unanimously concluded that it should be recommended to the Evaluation Board that any contract award should be made to Pathway”.
Sir Wyn Williams: Can I just check, Mr Beer, is the highlighting in these passages on the original document or subsequently?
Mr Beer: All original. We’ve not marked up any document at all. We have in our personal notes obviously but that is not what is being displayed. This is the state in which we received them.
Sir Wyn Williams: Right.
Mr Beer: That can be taken down, thank you.
Shortly thereafter, the Evaluation Board met under the chairmanship of Robert Peaple to discuss the findings of the Evaluation Team and to reach a decision on the selection of a supplier. For substantially the same reasons given by the Evaluation Team, the board reached the unanimous conclusion that if the project sponsors were to decide to award a contract, the award should be made to Pathway. The board reiterated that the selection of Pathway would “require a proactive management stance by sponsors, thereby acknowledging the identified weaknesses in areas such as the credibility of delivery by Pathway and weaknesses in Pathway’s management capability.”
We can see that, but needn’t turn it up, at POL00028450.
Robert Peaple will be giving evidence to the Inquiry in the coming weeks, and he will state that the approach adopted by the sponsors to the procurement was “extremely thorough” and that the evidence made available to him does not indicate that there were any flaws in the evaluation process. He says that in his statement at WITN04020100.
You, Chair, may wish nonetheless to explore why it was that Pathway’s solution, whose implementation was known to involve a higher level of technical risk than its competitors, was chosen by the Evaluation Board. Endorsement of the board’s recommendation was subsequently sought from the Programme Steering Committee which met on 1 May 1996 under the chairman of Peter Mathison, chief executive of the Benefits Agency. The committee unanimously endorsed the conclusion of the Evaluation Board that Pathway should be selected as the chosen supplier of the services. Although the committee observed that Pathway’s technical solution involved “an added element of technical risk”, the committee expressed confidence that Pathway’s performance could “be managed effectively” by the joint project management arrangements which had been agreed between Post Office Counters and the Benefits Agency. We will see a record of that recommendation at DWP00000029.
Within the Post Office, the proposal to select Pathway as the chosen provider was scrutinised by both the Major Projects Expenditure Committee (known as MAPEC) and the Post Office board. MAPEC (whose members comprised Richard Close, Jeremy Cope and John Roberts, chief executive of the Post Office Group) met on 30 April 1996 and noted that Pathway’s technical solution was “the least preferred” of the three and “presented a higher risk to delivering the programme” by reason of its complex design, which was unproven on a large scale deployment and involved bespoke code which would have to be modified.
The committee observed both short-term and long-term risk in selecting Pathway included: late delivery; the pressure to accept incomplete functionality; premature rollout, which would prove to be unreliable; and, lastly, a fragile software system. However, the committee concluded that such risks were “manageable and could be effectively mitigated by stronger technical assurance, rigorous user and system testing and appropriate contingency planning by the supplier.”
In the expectation that appropriate conditions would be imposed to give effect to these measures, the Committee endorsed the selection of Pathway. We can see a record of that, and it needn’t be turned up, at POL00028451.
You, Chair, may wish to consider whether the Committee’s expectations in relation to strengthening technical assurance of the system were realistic in view of the inherent constraints, which arose from the transfer of risk to the supplier, for the system’s design and development.
At a meeting of the Post Office Board on 7 May 1996, authority was granted to the Post Office Counter to enter into contracts with the Benefits Agency and Pathway, subject to the resolution of a funding issue affecting the Social Security Agency (Northern Ireland). Having approved the selection of Pathway and having obtained authority from the Treasury for funding of the contract, Peter Lilley, the then Secretary of State for Social Security, made a public announcement on 15 May 1996 confirming the selection of Pathway as the chosen contractor to finance, supply and operate the new automated system for Post Office transactions and benefit payments.
Can we look at that, please, at POL00089859.
The Secretary of State announcement says:
“The Secretary of State for Social Security has today announced that the joint program to bring technology to Post Offices and benefit payments, Benefits Agency and Post Office Counters Ltd has chosen Pathway as the contractor to finance, supply and operate the new automated system. This is a milestone of major significance to our organisations. Mr Lilley also announced the design of the benefit payment card which is integral to the new method of payment. The card will simply be known as the payment card.
“Pathway is a consortium led by ICL Plc, including such companies as De La Rue and Girobank. All consortium members are specialists in their own fields. They complement each other and ourselves by offering expertise in areas such as plastic card production and distribution and we shall now be starting a period of intense activity to further develop and test the new system.
“We expect the first benefit payment cards to be used in Great Britain by people in receipt of Child Benefit starting from the autumn.”
That’s autumn 1996.
“Although there has been some rescheduling of the timetable we still expect to have all post offices automated and all benefit payments, including those of the War Pensions Agency, being paid by card by 1999.
“For the Benefits Agency, changing the way benefits are paid at post offices will reduce benefit fraud because the computer system will confirm the amount to be paid to the customer and enhance the service to customers by providing a more secure and convenient alternative to giro cheques and order books. It will also support the Benefits Agency’s change programme, which aims to deliver our service more effectively and efficiently.
“For Post Office Counters Ltd, automation means greater choice, convenience and personal service to customers and clients, bringing it into line with the best retail practice. Automation will protect the nationwide network of Post Offices by securing a contract with the Benefits Agency for the next eight years. The introduction of this new technology will also help extend Post Office Counters Ltd’s product range and help it to retain and grow core and new business.
“We’ve come a long way since 1993 when Post Office Counters Limited and the Benefits Agency joined forces to introduce a new way of paying benefits at post offices. We still have a lot of work to do, but by continuing to work together we will remain on target to achieve the first payments by card at the first post offices later this year.”
Can I turn then to the second topic in Phase 2: design and development. Having been selected as the preferred candidate to bring modern technology to Post Office branches and benefits payments, Pathway entered into three related agreements. The first was with Post Office Counters, and the Department of Social Security acting jointly, and was known as “the Authorities Agreement”. We can see that – needn’t turn it up – at POL00028175.
The second agreement between Pathway was with Post Office Counters alone, and we can see that at POL00028181. And the third agreement was solely with the DSS.
We can see that at POL00028178.
Each of the three agreements specified the project sponsor’s functional requirements at a high level, leaving responsibility for the detailed design and development of the hardware and software necessary to deliver those services to Pathway itself.
There were essentially two key aspects to the automation project which Pathway undertook to deliver. Firstly, Pathway was obliged to devise a service which would enable the automation of benefit payments at Post Office Counters. The names given to the core element of this service were the Benefit Payment System, BPS, or Benefit Encashment System, BES, which were used interchangeably.
The purpose of BPS or BES was to enable those who ran Post Office branches to make payments to customers of the Benefits Agency upon presentation of their benefits payment card. The system was supported by additional backend systems developed exclusively for the Benefits Agency, which included a card management system and a payment authorisation system. In addition to the BPS, Pathway developed an additional optional service at the request of the Benefits Agency which was known as the Order Book Control Service, OBCS.
This provided a temporary mechanism designed to prevent fraudulent encashment of cancelled or stopped order books pending their complete replacement by the benefits payments card.
The second major aspect of the project related to the automation of other services and transactions at Post Office branches. This aspect of the solution was devised exclusively for Post Office Counters, and comprised of, firstly, an Electronic Point of Sale service, an EPOSS and, secondly, an Automated Payment Service, an APS. EPOSS was intended to process the purchase of Post Office products and services by customers of Post Office branches whilst the APS was designed to enable the incremental payment of bills by the customers of Post Offices, counter clients, such as utilities companies and the recharging of cards and keys used to access such facilities.
Part of the considerable complexity inherent in the project arose from the need to integrate these various services with parallel computing systems being developed internally by the Benefits Agency and Post Office Counters. One of the systems which Pathway was required to integrate was the Customer Accounting and Payment System, CAPS, which was being developed by the Benefits Agency in order to create a uniform method of recording and disseminating information relating to customer payments.
In order to ensure the accuracy of payments being made via the benefit card, it was necessary for data stored on CAPS to be fed to the systems which Pathway was developing to support the BPS.
The principal systems in the Post Office domain with which Pathway was required to integrate were firstly the Reference Data System which recorded details of, for example, Post Office branches, products and prices. Secondly, the Host Automated Payment System, HAPS, which managed Post Office Counter’s existing automated systems. Thirdly, Transaction Information Processing, TIP, which was required to extract data from Pathway’s systems in order to reconcile it with Post Office Counters’ own accounting and management information systems.
Shortly after the contract was awarded, ICL purchased the shares which De la Rue and Girobank owned in Pathway, and changed the company’s name to ICL Pathway Limited. And I will, from now on, refer to the relevant company as the “ICL Pathway”.
Responsibility for overseeing the development, delivery and integration of this complex web of integrated systems and services rested with something called the Programme Delivery Authority, PDA; which was staffed by representatives from both Post Office Counters and the Benefits Agency. The PDA was governed by a board whose membership was drawn from the senior management of ICL Pathway, from Post Office Counters, from the Benefits Agency, and from Social Security – the Social Security Agency of Northern Ireland.
A Programme Steering Committee provided direction and guidance to the programme. Its membership comprised Stuart Sweetman, the managing director of Post Office Counters; Peter Mathison, who, as I’ve said, was the chief executive of the Benefits Agency; Alec Wylie (spelt with a C), the chief executive of the Social Security Agency in Northern Ireland; and Keith Todd, the chief executive of ICL.
The first key milestone of the programme was the development and implementation of a pilot scheme known as Initial Go Live, IGL, the primary purpose of IGL was to demonstrate the successful functioning of the benefits payment card by the target date of 23 September 1996, scarcely four months after the award of the contract.
By reason of the very compressed timetable for delivery of the benefit payment system and its associated backend services, only very limited functionality was available. Therefore, the service was initially rolled out to a single Post Office in Stroud, Gloucestershire, and only permitted the payment of Child Benefit to 14 people. We can see that at FUJ00058278.
You’ll be hearing evidence from John Meagher, who acted as the Horizon product assurance manager. He describes this Initial Go Live as a “PR exercise” which “distracted from the primary objective and wasted valuable time and resources”. The cross-reference to that is WITN04150100.
You, sir, may consider that delivering the Initial Go Live by the September deadline was a Pyrrhic victory in the light of the chronic delays which subsequently beset the programme and to which I will turn, if it pleases you, after lunch.
Sir Wyn Williams: Impeccable timing. Thank you. 2.00, everyone. Thanks very much.
Mr Beer: Thank you, sir.
(12.58 pm)
(Luncheon Adjournment)
(2.00 pm)
Mr Beer: Sir, I hope by now, that you and Mr Page have got a live transcript –
Sir Wyn Williams: We have indeed.
Mr Beer: – on a screen in front of you.
Sir Wyn Williams: Yes, thanks very much.
Mr Beer: Before lunch I dealt with the Initial Go Live, namely rolling out to a single Post Office in Stroud of Child Benefit alone to 14 people.
The original timetable stipulated in the Authorities Agreement envisaged that an operational trial of all key services would take place between September 1996 and November 1997 and that rollout of both the benefit payment and Post Office Counter services would commence in July 1997. We can see that at POL00028173.
Although the Initial Go Live, such as it was, had been delivered on time, it soon became apparent that the original contractual timetable was unduly optimistic and by February 1997 the three contracting parties had entered into a “no-fault replan” of the program, which provided for phased implementation of the ICL Pathway services, with a view to completing the operational trial of the system by 21 November 1997, and commencing a national rollout to the Post Office Network on 24 November 1997. We can see that, no need to turn it up, at Fujitsu00097062 and Fujitsu00078186.
However, ongoing concerns about delays to the project prompted the three contracting parties to Commission a review of the program by PA Consulting Group, a management, information and technology consultancy, between August and September 1997.
The objectives of the review were to obtain an independent assessment of, firstly, the current weaknesses and risks in the program; secondly, the implications for its future delivery; and, thirdly, options available to guarantee delivery capability.
In a report delivered in October 1997, PA Consulting Group, concluded that ICL Pathway had “seriously misjudged” the scale of the development required to meet its contractual obligations and had, as a result, underestimated the resources and time required to deliver the technical solution, whilst recognising that delays had occurred in the CAPS program, which was being developed in parallel by the Benefits Agency, and that Post Office Counters needed to address deficiencies in its own resources, the PA Consulting Group concluded that the principal driver of further delay to the program was likely to be ICL Pathway and expressed doubt about the company’s ability to deliver the system for rollout by the summer of 1998, observing that ICL Pathway was proposing to defer known problems and system functionality to later software releases. We can see that, no need to turn it up, at POL00028092.
The precise causes of delay to the program remained a matter of dispute between the three parties. It was no secret that bugs, errors and defects were being detected during software testing. They were recorded by ICL Pathway in an incident management system known as PinICL, Pin-I-C-L in capital letters. Their existence was disclosed to the Programme Delivery Authority to enable an assessment of their business impact to be made and a decision taken as to whether efforts to resolve them could be deferred until a later software release. We can see that at FUJ00078989.
The mere fact that such errors were being detected during software testing was not generally thought to be a cause for significant concern, at least in the early stages of the software development. One issue of particular concern to the Programme Delivery Authority, however, was their ability to access ICL Pathway’s design documentation, in order to satisfy themselves that ICL Pathway was delivering a solution which would meet the project sponsors’ requirements.
Jeremy Folkes, who migrated to the PDA assurance team upon award of the contract to ICL Pathway and who will be giving evidence to the Inquiry in the coming weeks, will say that ICL Pathway senior management repeatedly refused to disclose documentation to the assurance team, citing as their reasons the terms of the private finance initiative upon which the party’s contractual arrangements had been based and the resultant transfer to ICL Pathway of the risks associated with design and development of the system.
Both Mr Folkes and John Meagher attribute this obstruction, at least in part, to Terry Austin, the ICL Pathway systems director, and we will see that in due course on Mr Folkes’s part at WITN05970100 at paragraph 77 to 79, and on Mr Meagher’s part at WITN04150100 at paragraph 15.
For its part, ICL Pathway’s stance, as articulated in a position paper, dated 5 March 1998, was that Post Office Counters and the Benefits Agency were improperly interfering in the design process and thereby causing significant delays to the programme. You can see that position paper at POL00031117.
Sir, you may wish to consider in due course whether ICL Pathway’s apparent reluctance to grant PDA access to design documentation reflected concerns about the quality of such documentation or, indeed, the absence of such documentation.
But whatever the true causes were of the considerable slippage in the programme timetable, it’s apparent from the evidence obtained by the Inquiry that the delay in rollout of the system had a very detrimental effect upon the finances of each of the contracting parties. We can see that, by way of example, at POL00028495.
The business case for the Department of Social Security had been predicated upon achieving substantial fraud savings which were not being realised during the prolonged development of the system. Moreover, the department was itself incurring additional administrative costs in funding the work of the PDA and the parallel CAPS program. ICL Pathway’s remuneration for designing and developing the system was deferred under the authorities agreement until the system had been rolled out, whereupon it was entitled to recover charges measured by reference to the volume of transactions processed by the system.
In the meantime, ICL Pathway faced a significant increase in its projected development costs by reason of the additional time and resources required to design and develop the system. Although Post Office Counters was arguably less exposed than the other contracting parties, it was also vulnerable to cost inflation, and was losing out on anticipated new sources of income.
You may wish to consider, in due course, what if any effect these financial pressures had upon later decision making as to the timing of the acceptance and rollout of the system.
As a result of the weaknesses in project management exposed by the PA Consulting Group review, the decision was taken to abolish the Programme Delivery Authority in April 1998 and transfer responsibility for managing the development and delivery of ICL Pathway services to a dedicated team within Post Office Counters, named Horizon. This team was led by David Miller, the newly appointed Horizon programme director.
One aspect of ICL Pathway’s solution, which was critically important to Post Office Counters, was the Electronic Point of Sale Service, the EPOSS, as I’ve already explained this element of the service was responsible for processing the transactions carried out within Post Office branches by customers purchasing products and services of the Post Office.
EPOSS represented the new automated mechanism by which subpostmasters would be required to account to the Post Office for the products and services sold in their branches. The reliability of the data process by EPOSS was therefore essential to enable the accurate reconciliation of the physical cash and stock held in the branch with the transactions performed by the subpostmaster at the counter. It would appear, however, that from a relatively early stage in the development process, ICL Pathway had itself concerns about the reliability of the EPOSS application.
You’ll be hearing from Terry Austin, systems director at ICL Pathway, who will state that he became aware of “a quality issue” with the early versions of EPOSS software, which have been developed using what’s described as a rapid application development technique which I understand to be an approach to software development that focuses more on ongoing software projects and user feedback and less on following a strict plan of development and testing cycles.
Mr Austin will explain that this agile but immature technique was rarely used in the development of large IT systems that had been developed by ICL Pathway, due to the very tight timescales for deliver of the service and the absence of a detailed business requirements specification.
Due to concerns about the robustness of the EPOSS application, ICL Pathway decided in July 1997 to instruct Escher, you’ll remember they were the small American company responsible for developing the Riposte platform, to reengineer the EPOSS code. According to Mr Austin, the EPOSS product was returned in November 1997, whereupon further functional changes and software fixes were applied to it by ICL path. We’ll see that in due course at WITN04190100 at paragraphs 22 to 23.
It would appear from the evidence obtained by the Inquiry that these measures did not resolve the problem of the EPOSS software’s underlying instability. As a result of ongoing concerns about the volume of bugs, errors and defects detected in the EPOSS publication. Terry Austin requested that a team of developers be established, known as the EPOSS PinICL taskforce, in order to identify the nature and cause of the outstanding defects in EPOSS and to fix them.
The taskforce operated between August and September 1998 with the principal objective of reducing to zero the number of PinICLs recorded against the EPOSS application.
In a report produced in September 1998 by Jan Holmes, ICL Pathway quality and audit manager, and Dave McDonnell, senior software developer, it was observed that “There were significant deficiencies in the EPOSS product, its code, and design”.
The cross-reference to that is FUJ00080690.
The authors of this report attributed this deficiency to “a combination of poor technical design, bad programming and ill thought out bug fixes”.
Moreover, they observed that “the negative impact of these factors will continue and spread as long as the PinICL fixing culture continues, partly due to the nature/size of the bug fixing task and partly due to the quality and professional of certain individuals within the team”. If we can just look at that please, it’s FUJ00080690. If we can turn to page 17, please and look at paragraph 7.3.
“Although parts of the EPOSS code are well written, significant sections are a combination of poor technical design, bad programming and ill-thought out bug fixes. The negative impact of these factors will continue and spread as long as the PinICL fixing culture continues. This is partly due to the nature/size of the bug-fixing task and partly due to the quality and professionalism of certain individuals within the team.”
Then some examples are given. The report warned that the application of what we’ll describe as fixes to the code, which had in 1997 been reengineered by Escher, created additional risks to the underlying stability of the product. The authors of the report expressed the fear that “code decay”, as it was called, which I understand to mean the tendency for computer software to be gradually worse in performance or responsiveness over time, eventually leading it to become completely faulty, unresponsive or unstable, and code decay might be caused because the code has been updated or altered in ways that have introduced more bugs and errors over time, especially in the light of the increasing difficulty or intricacy of the code modifications.
So the authors expressed fear that code decay will, assume it hasn’t happened already, cause the product to “become unstable” and they observed that “there is no guarantee that a PinICL fix or additional functionality can be made without adversely [affecting]” – as I think it the report was meant to read but it says “affect” – another part of the system.
We can see that earlier in the same document at page 7.
Under “EPOSS code” at the top there:
“It is clear that senior members of the Task Force are extremely concerned about the quality of code in the EPOSS product.”
Then reading on a sentence:
“Since then many hundreds of PinICL fixes have been applied to the code and the fear is that code decay will, assuming it hasn’t already, cause the product to become unstable. This presents a situation where there is no guarantee that a PinICL fix or additional functionality can be made without adversely [affecting] another part of the system.”
It is currently unclear whether the concerns that were being voiced internally by employees of ICL Pathway were, in the autumn of 1998, communicated to Post Office Counters or, indeed, appropriately addressed within ICL Pathway by the senior management of the company.
You may wish to consider in due course whether there was a failure by the senior management of ICL Pathway to be candid with Post Office Counters about the nature and scale of the problems which it encountered during the development of the EPOSS product and, if so, what if any effect this had upon Post Office Counters’ later decision to proceed with the project in spite of the Benefits Agency’s decision to withdraw from it.
Whether or not ICL Pathway were concealing from Post Office Counters the true extent of any issues they were experiencing with EPOSS, it is apparent that senior managers within Post Office Counters were aware of some concerns about the development of the EPOSS application. We can see some examples of that, no need to turn them up now, at POL00028587 and POL00028323. During Phase 3 of the Inquiry, you will be hearing evidence from David Smith, the head of automation transformation, who will state that:
“It was known that work on the desktop started in London but was then sent across to the US before being sent back to the UK for completion. This created some suspicion that Pathway weren’t up to the task.”
The reference for that is WITN05290100 at paragraph 35.
Mr Smith will explain that he later commissioned a report from a consultant employed by French Thornton, as they were then called, who was briefly granted access to ICL Pathway to details of the EPOSS design. The consultant reported back that ICL Pathway had not adopted a modular design in the development of EPOSS but rather a method by which all transactions were routed through a core Horizon. The consequence of adopting this approach was that when a new transaction was added he said, it would, through its core, be linked to other transactions and it could introduce error into pre-existing transactions.
The senior management at Post Office Counters were advised that the problems posed by this method of design could not easily be remedied and the adoption of a new approach would require the EPOSS application to be rewritten. The cross-reference to that is WITN05290100, at paragraphs 41 to 45.
By reason of continued slippage in the project timetable, Post Office Counters and the Benefits Agency had, by early 1998, begun to prepare for potential litigation with ICL Pathway, in the event that the Horizon System could not be delivered to a satisfactory state or within a reasonable timescale.
In order better to understand where responsibility lay for ICL Pathway’s failure to complete the operational trial by the revised deadline of November 1997 and the overall strengths and weaknesses of their case, Post Office Counters and the Benefits Agency instructed a firm called Project Monitors (sic), who were an independent firm of consultants, to provide expert assistance.
Having initially conducted a number of investigations into the causes of delay to the program, Project Mentors were instructed to carry out an assessment of the requirements analysis undertaken by Pathway for the benefit payment service. In carrying out that analysis, Project Mentors did not have access to ICL Pathway’s own internal documentation and instead relied upon documents in the possession of the project sponsors, which had either been disclosed to them or produced by ICL Pathway the cross-reference for that is POL00031114.
In a report dated 18 December 1998, Project Mentors found that ICL Pathway had “made no attempt to undertake requirements analysis in accordance with normal industry practice”, despite having access to the sponsors’ requirements for the benefit payments service at the time of entering into the contract.
As a result of this failure, the sponsor’s expert concluded that ICL Pathway had designed and partially built a system without knowing whether it fully met the requirements of the Department of Social Security.
Project Mentors acknowledged in their report that the analysis which they had undertaken was restricted to the benefit payment service. However, they expressed “grave concerns” that the same lack of professional analysis would be apparent in other areas of the system, and articulated a particular concern about the EPOSS application. The cross-reference to that is POL00031114 at paragraphs 2.2 and 2.3.4.
Reflecting on the impact that this failure was likely to have upon the future of the programme, Project Mentors made an important observation. I wonder whether we could look at POL00031114 and in particular page 7.
Try page 11, please. Then one more page, please. Thank you.
Perfect.
Under its conclusion “Impacts on the Programme in the Future”, Project Mentors said:
“Our experience of systems where requirements have not been analysed satisfactorily is that the system fails to meet the users’ needs. An effective acceptance test will identify many such failings necessitating considerable rework. The result is a significant extension of the time and cost required to complete the system and roll it out. The alternative is to allow unacceptable processing in the operational environment, with unpredictable and potentially damaging results.”
“In our opinion the failure to satisfactorily analyse the requirements for the Benefits Payment System makes it unlikely that the users’ needs will be met by the current Pathway system.”
We can take that down. The findings of this report were circulated to senior managers within the Benefits Agency and also within Post Office Counters, including David Miller, the Horizon’s programme director, who will be giving evidence during Phase 2 of this Inquiry. The evidence of distribution is given in POL00038829.
It’s presently unclear what action was taken by Post Office Counters, and the Benefits Agency, to address the concerns which had been articulated by their independent expert, Project Mentors, in December 1998. You, sir, may wish to explore what, if any efforts, were made by those who had knowledge of these concerns, to bring them to the attention of senior officials and ministers within Government, who were at that stage charged with making decisions about the future of the Horizon project.
Can I turn, then to the third topic in the Phase 2 opening, namely the Treasury review and cancellation of the benefit payment card.
I’d like to begin by saying a few words about the relationship that existed at this time between the Post Office and the Government. By the late 1960s, the Post Office had ceased to be run as a Department of State and had instead been transformed into a statutory corporation. Its governing body consisted of a chairman and board members, each of whom were appointed by the Minister of Posts and Telecommunications, subsequently the Secretary of State for the Department of Trade and Industry. See section 6 of the Post Office Act 1969.
The day-to-day operations of the Post Office were largely managed independently of government. However, the Secretary of State was empowered to give directions to the Post Office, with which the chairman and board were applied to comply. See section 11 of the Post Office Act 1969.
As its sponsoring department, the DTI exercised oversight of the Post Office thorough its Postal Directorate and was kept abreast of significant developments in the programme to automate Post Office Counters and the payment of benefits.
As I’ve already explained, ICL Pathway had failed to meet the revised contractual deadline to complete an operational live trial of the new Horizon System by 21 November 1997. As a result of this default, the Programme Delivery Authority wrote to John Bennett, managing director of ICL Pathway, on 24 November 1997 formally notifying him, on behalf of both the DSS and Post Office Counters, that ICL Pathway was considered to have breached its contractual obligations. We can see that at POL00028442.
ICL Pathway disputed that it was in breach of contract, and attributed the delays to a variety of factors, including: (i) insufficient specification in the sponsors’ requirements; (ii) later attempts by the sponsors to enhance those requirements; (iii) improper interference by the sponsors in the design of the system; and (iv) delays in the CAPS program. We can see that response from ICL Pathway at POL00031117.
By early February 1998 the Department of Social Security was sufficiently concerned about the state of the project that Harriet Harman, then Secretary of State, wrote to Tony Blair, the then Prime Minister, to notify him that there was “a serious risk” that the project would fail either to deliver its objectives at all or to do so within a worthwhile timescale. I wonder whether we can look at that, please, at CBO00000041.
Ah, perfect.
Thank you very much. This is the letter from Harriet Harman to Tony Blair.
“Prime Minister …
“I am writing to alert you to problems with a private/public sector project designed to improve the benefit payment system and automate the Post Office network. In a nutshell, there is a serious risk this project will fail to deliver its objectives – or will not do so within a timescale that will make it worthwhile.
“We have inherited this £1 billion project from our predecessors. Understanding the PFI programme, ICL were contracted by my Department jointly with Post Office Counters … to fund the development of a system which will provide an automated means of paying benefits through a Benefit Payment Card to replace our current expensive and insecure paper-based methods; and, on the back of this, to provide some additional technology to Post Offices, with the aim of helping [Post Office Counters Limited] to extend the range of their future business. The project was supposed to be fully operational in all Post Offices by the end of this year. But due to a long history of problems, ICL have been unable to deliver to this timetable, so the completion date has been extended by two years to 2000 at the earliest, with a consequential loss to us of around £200 million in fraud savings foregone.
“This two-year delay also brings with it funding consequences both for ourselves and the private contractor. These and some continuing major concerns about the project’s deliverability have led us to the view it would be prudent to take stock before committing to further investment. In doing so, it must be right to look at the medium and longer-term Government objectives for a modern welfare system and the best way to sustain a rural Post Office network into the millennium and beyond.
“The questions this will raise have far-reaching consequences, which could impact on our future capacity to deliver the benefit payment service people want, and on the future of the Post Office network, vital to so many of our communities. You will anyway wish to be aware because of the wide-ranging Ministerial interests involved, and the need to reach a cross-Government consensus on the best route forward. The primary Departmental responsibilities fall to Margaret Beckett and me, but there are also significant issues around rural policy, the Government IT infrastructure and strategies to promote social cohesion.
“It is now urgent to reach a decision on the best way forward, because further investment will be required if the project is to continue. It is also possible that ICL, the private contractor concerned, may give their version of the story to the press at any moment, in an attempt to influence of the outcome. There are inherent dangers of damaging media speculation, especially around the future of Post Office: as you know, the Post Office involvement means there could be matters here arousing strong public interest – and opinions. There is already some speculation in the specialist Press.
“You may also wish to be kept informed because ICL are part of the Fujitsu empire, with whom you have had some contact on your recent visit to Japan and there is also a constituency dimension for you.”
More about that later:
“The immediate urgency on the financing of the project means that John Denham (who is leading this work here) will need to write very shortly to Alistair Darling on the funding issues. Although we have a clear view about the best route forward, I am clear that we should at all costs seek to avoid getting focused too closely on individual Departmental interests, and that we should aim to establish an agreed view from the start, avoiding the potentially damaging effects that our quite legitimately differing Departmental views can have, when exposed to the public eye. On this basis, I shall be seeking to discuss with Alistair and Margaret in the near future.”
Ms Harman’s letter to the Prime Minister appears to have prompted the intervention of Alistair Darling, the Chief Secretary to the Treasury and triggered the establishment, in early March 1998, of the interdepartmental Horizon project review group. The group comprised of senior officials drawn from the Treasury, the Department of Trade and Industry, the Department of Social Security, and the leadership of the group was provided by senior treasury officials, as a result of which the work undertaken by the group came to be known as “the Treasury review”.
During its inaugural meeting, the group identified two distinct stages to the review: the first was a project assessment intended to evaluate the viability and likely cost of continuing with the Horizon programme. The second stage involved contingency planning to establish the alternative options open to government in the event that the decision was taken to cancel the project. We can see that planning at CBO00000017 and BEIS0000104.
Responsibility for the project assessment was entrusted to three individuals who collectively became known as “The Treasury review panel”. The panel was chaired by Adrian Montague, chief executive of the Treasury PFI taskforce, and also comprised Bill Robins, Director General of Communications and Information Services in the Ministry of Defence, and Alec Wylie, chief executive of the Social Security Agency for Northern Ireland.
PA Consulting Group, the consultancy firm which had earlier produced a report back in October 1997 into the causes of the project delays, was also instructed to support the work of the panel by conducting an assessment of the technical viability of the Horizon programme.
In a report dated July 1998 the Treasury review panel concluded that the programme was “technically viable”, meaning, I think, viable from a technological point of view and perspective, as opposed to theoretically viable. Whilst recognising that there were some risks relating to the scalability and robustness, given that the system had been tested only at the level of its component parts, the panel stated that it was satisfied that these risks were being “well managed” by ICL Pathway. The panel also observed that “The basic infrastructure is very robust for the future”, and in the main, industry standard products have been used. The cross-reference to that is POL00028094.
Adrian Montague, the chair of the panel, who will be giving evidence to the Inquiry, will state “None of the shortcomings of the Horizon System that came to dog the project in later years were or could have been evident at the stage that the panel made its assessment”.
He explains, amongst other things, that much of the system’s functionality was still under development at that stage and it wasn’t possible at that stage to undertake acceptance testing of the system in the live operation.
The cross-reference to that is WITN04000100, at paragraph 11.
You will also be hearing from Peter Copping, a director of PA Consulting Group, who conducted the technical aspects of the project’s assessment on behalf of the panel. He will explain that he carried out several investigations for the panel involving meetings and one-to-one discussions with ICL Pathway, the Benefits Agency and Post Office Counters.
The cross-reference to that is WITN03970100. As you will recall, ICL Pathway was sufficiently concerned by this time about the robustness of the EPOSS application, that it would, within a matter of weeks, establish a dedicated taskforce to address the significant volume of suspected errors, bugs and defects in the EPOSS system. The cross-reference is Fujitsu00080690.
Given the extent of the concern within ICL Pathway about the reliability of the EPOSS application, you, sir, may wish to explore the extent to which these matters were brought to the attention of the Treasury review panel and how, if at all, they affected the panel’s assessment of the developing the systems fitness for purpose.
The Treasury Review Panel ultimately concluded that termination of the entire project was not a viable option on the grounds that it would further delay the automation of Post Office Counters leading to the likely loss of future income and would cause significant reputational and financial harm to ICL Pathway.
In the alternative, the panel proposed a restructuring of the programme, in which either the benefit card would be rolled out and transfer to payment by ACT would be deferred until Post Office Counters’ infrastructure had adapted to support banking, financial services and other applications, with an agreed contract extension, which would reduce the overall loss to ICL Pathway – this was described by the panel as option 1 – or (b) the benefit payment card would be abandoned and the paper-based method of paying benefits would continue until Post Office Counters was ready to complete for delivery by ACT with compensation being made to ICL Pathway for the reduction in the scope of its contract. This was described by the panel as option 2.
The cross-references there are POL00028094 in annex A.
Upon receipt of the panel’s recommendations, the Horizon Project Review Group, which had now been renamed as the Horizon Working Group, produced is own report in which the divisions within Government over the future of the project were very plainly exposed. Neither the DSS nor the social security team within the Treasury were in favour of pursuing option 1, on the grounds that it would postpone full migration to payment via ACT by up to five years, it would involve two significant changes in the method of payment to benefit recipients over the space of a few years and it would leave the DSS exposed to the costs of further delay. The cross-reference there is HMT00000034 at paragraph 4.3.2.
The DSS favoured cancellation of the benefit payment card, whereas the social security team within the Treasury favoured terminating the entire project, we can see that in the same reference at paragraph 1.7. By contrast, the remaining members of the Working Group, including the Department for Trade and Industry, were in favour of proceeding with the benefit payment card, subject to agreeing suitable terms with ICL Pathway. They envisaged that a contract extension would be needed but stipulated that there should be no increase in the level of remuneration paid to ICL Pathway. They said that, if ICL Pathway was not prepared to accept these terms or the negotiation of terms for a contract extension were unsuccessful, it was recommended that Post Office Counters and the DSS should terminate the contract on the grounds of ICL Pathway’s non-performance. The same reference, paragraph 1.7.
The report produced by the Horizon Working Group requested “an early steer” from ministers, but acknowledged that there were a number of competing considerations and that any litigation with ICL Pathway would “inevitably be messy and uncertain”. Same reference, paragraph 1.5.
Ministers ultimately sanctioned a time limited period of negotiation with ICL Pathway with a view to agreeing suitable terms for the continuation of the project in its original guise. Stephen Byers, then Chief Secretary to the Treasury appointed Graham Corbett, deputy chairman of the Monopolies and Mergers Commission to act as a “troubleshooter” and to facilitate the negotiations.
By October 1998, Graham Corbett reported back to the chief secretary that the parties had been unable to reach agreement on a commercial basis for continuing the full project.
As a precondition for delivering the contractual services, ICL Pathway had demanded a positive return on its investment. This position was unacceptable to ministers, as it sought to reverse the contractual allocation of risk, which Pathway had agreed in the spring of 1996 and was therefore accordingly reject. The cross-references to that report are BEIS0000284 and POL00028098.
Further attempts to salvage the project were made by Post Office Counters and ICL Pathway in late autumn 1998. As part of those ongoing negotiations, ICL Pathway proposed changes to the contractual terms for the acceptance of the system. I shall address those contractual terms a little later in the opening. However, for present purposes, it’s sufficient to note that ICL Pathway’s proposals envisaged the system being accepted, and the project sponsor’s rights of termination being lost, upon the conclusion of technical and model office tests, rather than an operational live trial.
Moreover, ICL Pathway proposed a relaxation in the criteria for acceptance, which would have resulted in the system being rolled out nationwide with a much higher incidence of unresolved technical defects. The cross-reference there is POL00028679.
Whilst these proposals were under consideration, the results from ICL Pathway’s latest round of software testing were being assessed by Mr David Miller, the programme director for Horizon, and found to reveal “Significant problems with the way Horizon passes information through to TIP”, that is transaction information processing.
We can see that in his memo dated 16 November 1998. I wonder whether we could turn that up, it’s POL00028421. You can see the date of 16 November 1998, the distribution list at the top, and in the second paragraph, Mr Miller says:
“We are due to start the Model Office Test proper and the final pass of End to End on 14 December 1998. My present assessment is that there are some significant problems with the way Horizon passes information through to TIP. These relate to the provision of balanced outlet cash accounts and the processing of the ensuing information via TIP. Whilst we allocated some extra time during the Corbett review to sort out any outstanding issues, we need to be aware of a potential threat to the 14 December date.”
As I explained in my introduction to the design and development of Horizon, TIP (transaction information processing) was a system development by Post Office Counters to obtain cash account and transaction data from Horizon, in order to feed that data back into Post Office Counters’ existing accounting systems. It is evident from Mr Miller’s memo that these results were of significant concern to Post Office Counters, and threatened to cause delay, further delay, to the software testing cycles, which needed to be completed, prior to entry into the operational live trial of the system.
In the light of the poor results which had emerged from the software testing in the autumn of 1998, you, sir, may wish to consider whether ICL Pathway’s proposal to make significant changes to the conditions for acceptance reflected, in fact, a lack of confidence in the integrity and robustness of their own solution.
That document can come down, please.
One of the principal reasons why the Treasury Review Panel had discounted the option of terminating the entire contract in July 1998 was the harm which ICL would have suffered to its reputation and finances. By October 1998, ICL was projected to sustain losses of more than £200 million, in the event that the project was cancelled. We can see that from POL00028098. Losses of that magnitude were not only a cause of serious concern to ICL but also to Fujitsu Japan, which by this stage had acquired all of the company’s shareholding.
In early December 1998, Michio Naruto, the vice chairman of Fujitsu and chairman of ICL, contacted Sir David Wright, the British ambassador to Japan, to articulate Fujitsu’s concerns about the progress of negotiations over the future of the Horizon project. Mr Naruto warned the British ambassador that the failure of the project would have “serious repercussions for Fujitsu’s international standing, lead to major internal difficulties with Fujitsu, and the collapse of ICL”. The cross-reference there is BEIS0000336 and 7.
In a subsequent report sent to the Cabinet Office, the Treasury, the DTI and the DSS, dated 4 December 1998, Sir David Wright, the ambassador, expressed the view that Fujitsu were not making idle threats about the future viability of ICL which, he said, if permitted to fail, would “have profound implications for jobs in the UK and for bilateral ties”.
He observed that the failure of the Horizon project would “be damaging politically, at home, and to the UK’s position of strength in Japan as against its European competitors”. The cross-reference there is again BEIS0000336.
Within a matter of days, the question of the project’s future was escalated to the then Prime Minister, Tony Blair. I wonder whether we could look, please, at CVO001401.
There seems to be a problem with that; evidence that not all IT projects move smoothly. Although it’s early, this is a relatively important document and I just ask for ten minutes, if we take our break early, and I’ll seek to ensure that’s uploaded to the system and viewable by everyone concerned.
Sir Wyn Williams: Excellent. Right. Ten minutes.
(2.52 pm)
(A short break)
(3.02 pm)
Mr Beer: Thank you for the time you allowed me.
You’ll remember that I had just spoken about communications from Michio Naruto to Sir David Wright, the British ambassador in Japan, and then the escalation of those concerns to the Cabinet Office, to the Treasury and the DTI, and then the reference made by Geoff Mulgan to the Prime Minister. Can we turn up, please, CBO0010001.
This is the minute or memorandum about which I was speaking before the break, from Geoff Mulgan, a senior civil servant in the Cabinet Office and a member of the Horizon Working Group, as it had then been called.
Sir Wyn Williams: Could I just have the reference number again, it’s gone off the top of the –
Mr Beer: CBO00100001, “CBO” indicating that this was disclosed by the Cabinet Office.
Sir Wyn Williams: Yes, I’ve got it yes.
Mr Beer: You’ll see it’s addressed to the Prime Minister in the copy list on the top right-hand side. In paragraph 1, Mr Mulgan says:
“A decision now needs to be taken on whether to proceed with the Horizon project. You will recall this is the initiative to automate the Post Office network, involving ICL. The project is nearly three years behind schedule, having been plagued with problems. Negotiations have been underway since the summer to find a way to continue the project.”
Then over the page at paragraph 4:
“However, the decision is not clear-cut. The problems that have beset this project may well continue; continuation would lock the government in for 10 to 12 years to what may many see as a flawed system; cancellation on the other hand would enable the Post Office to take advantage of newer, cheaper and more flexible technology, while the DSS could move rapidly to paying benefits into people’s bank accounts. Cancellation would also release about [£2 to £3 billion] over the next decade to be spent in other ways to support and automate the Post Office.”
Paragraph 6:
“In making a judgement, the following issues are paramount …”
First:
“The virtues of the project itself: overall, Horizon now looks increasingly flawed. It is centred around a technology, the Benefit Payment Card (BPC), that’s both overengineered – and very expensive – and likely soon to be obsolete. Indeed, ICL acknowledge that the BPC will have no commercial value to them at the end of the project. Although they remain underdeveloped, the alternatives, which involve simpler off-the-shelf banking technology, look increasingly attractive, offering a route to universal banking, automated Post Offices and better provision of government information.”
Secondly:
“The effects on the Post Office network: cancellation would undoubtedly be destabilising. Subpostmasters fear that without Horizon they will lose their customer base. Their concerns can be partially, but not wholly, addressed through commitments on subsidy and promises that there will be government support for an alternative automation system. Rural subpostmasters’ fears will be particularly acute. Subsidies could be structured to keep the great majority of these open but, whatever happens, there will be a continuing stream of closures.”
Third:
“Effects on ICL: Cancellation would directly affect 270 jobs in ICL and more in suppliers. It would destroy ICL’s prospects of flotation and lead to a heavy (£200 million) loss this year, although even if the project continues, the £250 million losses incurred so far may at some point may have to be written off (NB the figures in the attached paper are wrong on this issue). Cancellation would also have a big effect on their reputation. It would probably force the resignation of the Chief Executive and put them into a protracted legal battle with government, in which they would seek to pin the blame on government in general, and DSS in particular.”
Fourth:
“Effects on Fujitsu: Fujitsu have provided the financial commitment to secure a deal. This is a sign of how important the project is for them. There would undoubtedly be a cost in terms of UK relationships with them in particular, and possibly with Japanese investors in general. They would argue that the government was guilty of a breach of faith: blaming ICL for what is in effect a change of policy.”
Fifth:
“Effects on PFI: continuation would set a precedent for renegotiating PFIs that are failing; cancellation would make funding of large IT projects harder in the future.”
Then, finally, over the page, the sixth point:
“Effects on Benefits Agency and welfare reform: continuation would hold up ACT. DSS estimate that they would save £800 million if the project was cancelled. Government would lose the opportunity to bank the unbanked.”
In paragraph 7, Mr Mulgan observed that government departments remained divided on the project:
“Alistair Darling remains strongly opposed to continuing. Ian McCartney for DTI will argue strongly for accepting a deal (Peter Mandelson has largely kept out of the discussions). The Treasury is divided at an official level, but Stephen Byers will probably, on balance, want to accept the deal for pragmatic reasons, even though he would prefer to cancel.”
At paragraph 8 he said:
“At first glance, most of these factors point towards continuation. However my view, which Lord Falconer broadly shares, is that although short-term considerations and expenditure point strongly towards making a deal, this will in the long-run prove unsatisfactory, leaving the Post Office and government dependent on a hugely expensive, inflexible, inappropriate and possibly unreliable system.
“In our view, the best outcome would be a deal with ICL to continue with an automation strategy which drops the Benefit Payment Card but focuses instead on helping the Post Office to provide banking services. This would enable a rapid move towards ACT and Post Office automation, but without the many problems associated with the [Benefit Payment Card]. It would enable good relations to be maintained with ICL and Fujitsu. The costs to government and ICL resulting from the [Benefit Payment Card] could legitimately be blamed on the previous government.
“However, it would not be possible to reach this position without first rejecting ICL’s offer and making it clear that the government no longer wishes to continue with Horizon in its present form. There would then be a danger that ICL would refuse to negotiate over a deal which included the [Benefit Payment Card]. In other words, this option is only feasible if the government is prepared to accept the significant risk that it would in practice lead to complete cancellation.
“Our view is that the case for cancellation is, just, strong enough to warrant that risk, although it would need to be announced alongside a strong package of support for Post Offices and for an alternative automation strategy …
“There are essentially three options:
“Option 1: Attempt to conclude a deal, broadly along the lines offered by ICL.
“Option 2: Seek a deal with ICL but excluding the Benefit Payment Card.
“Option 3: Reject ICL’s offer, move to terminate the offer and press forward with an alternative.
“Do you wish to give a steer?”
Then the handwritten addition appears to say:
“I would favour option 1 but for GM’s statement [perhaps] that the system itself is flawed. Surely there must be a clear view on this. Speak to me on [Monday], ie reading the enclosed paper, it all … on the financial deal. But the risks are pretty even, probably coming down on the side of continuing. The real heart of it is the system itself.”
We can get some further light on that on the next document that I would like you to look at, please, which is CBO00000009, thank you.
This is the formal response to Geoff Mulgan’s submission coming from the late Jeremy Heywood, the Prime Minister’s Principle Private Secretary, on 14 December 1998 in the following terms. You’ll see it’s addressed to Mr Mulgan from the Prime Minister’s Principle Private Secretary, Jeremy Heywood:
“The Prime Minister was grateful for your minute of 9 December.
“The Prime Minister was concerned about your view that the Benefit Payment Card is over-engineered and is likely soon to be obsolete. His clear preference would be to avoid cancelling the project, but to go for a variant of your Option 1 and 2. We should retain the Benefit Payment Card but seek to ensure that over time it delivers real benefits and provides an effective transition path to a satisfactory long-term position. If necessary the Prime Minister thinks it may be sensible to give ICL a financial incentive to improve the [Benefit Payment Card] project in this way.”
That can come down, thank you. After further interdepartmental negotiations reached an impasse in late December 1998, the Prime Minister was again invited to make a decision on the future of the project. Again, this was communicated through his Principal Private Secretary, Jeremy Heywood, who again wrote to Geoff Mulgan, this time on 6 January 1999. Can we look, please, at CBO00000009.
You’ll see the addressee and the author, Jeremy Heywood:
“The Prime Minister was grateful for your minute of 24 December.”
I’ve skipped over that in the interests of some speed:
“The Prime Minister thinks that we should proceed as follows:
“(i) we should indicate to ICL as soon as possible that we want to have further discussions with them on how the BPC project can be developed to facilitate the earliest possible move to smart cards.
“(ii) but we should also make clear to them that, if agreement cannot be reached within a specified time period on a development of the current project along these lines, the Government will stick with the current project.
“Notwithstanding (ii), the Prime Minister is clear that we should seek to move as quickly as possible to paying benefits directly into bank accounts via a smartcard facility that can be used in Post Offices.”
You will be hearing from Ian McCartney, Minister of State in the Department of Trade and Industry, who will say that these decisions of the Prime Minister to “take cancellation off the table” was an important moment. It was an important moment, he says, in the cross-governmental debate over the future of the Horizon project. The cross-reference for that is WITN03770100.
The Prime Minister’s intervention did not, however, lead to an immediate resolution of the stalemate within government, as the DSS remained firmly opposed to the continuation of the Benefit Payment Card. Alistair Darling will give evidence to the Inquiry about the opinion which he forms of the project, upon being appointed Secretary of State for the DSS in July 1998.
You will hear his view that the Benefit Payment Card, he considered, whether it was the right throughout the period of negotiation, was the right way to achieve the reform of the Benefit Payment System.
As the negotiations with ICL Pathway dragged on into the spring of 1999, further pressure was brought to bear on the UK Government by Fujitsu. Tadashi Sekizawa, chairman of Fujitsu, wrote to the Prime Minister in a without prejudice letter of 7 April 1999, claiming that Fujitsu would be required to make a provision in its accounts of approximately £300 million, unless a legally binding agreement for the future of Horizon was in place by 23 April 1999.
Mr Sekizawa warned the Prime Minister that neither ICL nor Fujitsu would be prepared to make any further investment in the development of the Horizon System if Fujitsu were obliged to make such a provision in its accounts, and expressed a hope that a resolution of the commercial negotiations could be reached. That’s CBO00000033.
Mr Sekizawa’s letter was followed by a meeting between the Prime Minister himself and Mr Michio Naruto, as I’ve said, the then Vice-Chairman of Fujitsu and chairman of ICL Plc, on 12 April 1999, in which a discussion took place relating to the future of the Horizon project, so within five days of the receipt of the Fujitsu letter. The cross-reference for that is CBO00000059, at page 5.
You’ll be hearing from Steve Robson, a senior civil servant in the Treasury, who was instructed to lead on the negotiations with ICL in early 1999 and who attended the meeting between the Prime Minister and Mr Naruto on 12 April 1999. He will state that the Prime Minister gave no commitment to Mr Naruto at that meeting.
You will wish to consider what, if any, influence, the pressure brought to bear by Fujitsu had upon the outcome of cross governmental negotiations about the future of the project.
On 11 May 1999, Jeremy Heywood the Principal Private Secretary to the PM, wrote to the Chief Secretary to set out the Prime Minister’s requirements for a resolution of the prolonged negotiations. The cross-reference is CBO00000053. The letter explained that:
“Any solution should meet three political requirements.
“(i) we do not want a huge political row with the Post Office or the subpostmasters lobby claiming that the entire rural network has been put in danger by the government.
“(ii) we should not put ICL’s whole future at risk.
“(iii) it would be important to ensure that the Government had a fully defensible position vis à vis the Public Accounts Committee.”
By this stage, the proposal which commanded the greatest cross governmental support envisaged the cancellation of the Benefit Payment Card and the purchase by Post Office Counters of the Horizon platform, including the EPOSS and to OBCS systems. This proposal did not however find favour with Post Office Counters or the Post Office Board, with the latter, the Post Office Board, only agreeing reluctantly to sign heads of agreement with ICL Pathway to continue the project without the BPC, the cross-references to that are POL00028612 and POL00028609.
It’s currently unclear to the Inquiry precisely what those responsible for deciding on the future of the project knew about the technical integrity and the robustness of the system. However, in a briefing from the Treasury to the Prime Minister’s office, dated 23 April 1999, there can be found a document entitled “ICL Pathway: list of failures”, in which the following is recorded.
Can we look, please, at CBO00000058 and at page 11. Whether we could expand on that a little bit, thank you. Just to explain, sir, some of these documents are in a better format than others because some of them have been retrieved by photographing at The National Archives the document, rather than being able to take them up and scan them. This is one in the former category.
So “ICL Pathway: list of failures”, this a briefing from the Treasury to the Prime Minister:
“Independent reviews of the Horizon project by external IT experts have all concluded (most recently this week) that ICL Pathway have failed and are failing to meet good industry practice in taking this project forward, both in their software development work and in their management of the process.
“To date in the development stages of the project:
“All planned release dates have been missed, including the key contractual milestone for completion of the operational trial, which ICL Pathway were placed in breach in November 1997 …”
Skipping the next one:
“Every release has been subject to reductions in the originally plant functionality.”
Next bullet point:
“And even when each release has gone live, there have been faults and problems which have resulted in the need for Pathway to reimburse DSS.”
Next bullet point:
“In the current trials, the known problems have risen from 46 in November 1998 to 139 at the end of March 1999; and currently 146 have not been resolved.”
It is therefore clear that some senior officials within the Government were aware of technical issues with Horizon, which called into question the integrity and robustness of Horizon. You, sir, will wish to explore how widespread this knowledge was within Government and then what was done with that knowledge.
Can I turn to the fourth and final stage of my opening in relation to Phase 2, acceptance and rollout. That document can come down. Thank you.
This concerns the acceptance of the Horizon System by Post Office Counters and the decision to commence rollout of the system nationwide in January 2000.
I shall begin what I mean by explaining acceptance of the Horizon System. As I explained earlier, the detailed design and development of the Horizon System was entrusted under the terms of the related agreements to ICL Pathway, who bore the risk that the system which it produced would ultimately fail to meet the requirements of Post Office Counters and the Benefits Agency, and would consequentially be rejected.
Responsibility for determining whether the system that had been development by ICL Pathway in fact met their requirements rested with Post Office Counters and the Benefits Agency. However, their right to terminate the agreement for failure of the system to comply with these requirements was narrowly subscribed in schedule AO7 of the authorities agreement. There is no need to turn it up, but the cross-reference is POL00028163.
The right to terminate was narrowly circumscribed as follows:
“The authority’s right of termination for failed acceptance shall only apply … in the event that (a) one or more high severity deficiencies, or (b) 10 or more medium severity deficiences remain uncorrected at the end of the operational trial period.”
The effect of this provision was that Post Office Counters and the Benefits Agency would have been legally obliged to accept the system if, on completion of the operational live trial, the conditions for acceptance had been satisfied, namely that there existed no outstanding high severity deficiencies and less than ten outstanding medium severity deficiencies within the system.
By the spring of 1999, one of the key areas of dispute which emerged between Post Office Counters and the Benefits Agency, concerned the readiness of the Horizon System to enter the operational live trial.
Between February and March 1999, ICL Pathway had for the first time conducted model office and end-to-end testing of all aspects of the technical solution, including the interfaces with Post Office’s backend systems, the cross-reference there is FUJ00058445.
The concerns articulated by the Benefits Agency in March and April 1999 appear to have centred upon four key areas: (i) the number and severity of outstanding faults identified in the system during software testing, (ii) the risk that significant numbers of new critical errors might emerge during the operational live trial of the system; (iii) the ability of further software testing to address ongoing concerns about systems stability; and (iv) the risks that new software changes introduced to the operational live trial might undermine the integrity of the system.
The cross-references are POL00028407 and POL00028406. As a result of those concerns, the Benefits Agency refused in late April 1999 to grant approval for entry into an operational live trial, insisting instead that a further cycle of model office and end-to-end testing be undertaken. That’s POL00043592 and POL00028405.
You’ll be hearing from Andrew Simpkins, a management consultant employed by Post Office Counters, to support the management of software testing and release authorisation.
He will state that Post Office Counters considered that further cycle of model office and end-to-end testing would be “of limited benefit when compared to the value to be gained from monitoring and evaluating the experience of actual users within a limited live trial”. That’s WITN06090100, paragraph 12.
Sir, you may wish to consider in due course what efforts were made by Post Office Counters to obtain feedback from end users during the live trial and the early rollout of the system and whether these were adequate to identify and address the challenges that such end users faced.
The cancellation of the benefit cards payments ultimately was to resolve this impasse and enabled Post Office Counters and ICL Pathway to enter an operational live trial of the Horizon System in late May 1999. By this stage, the scope of the services which ICL Pathway was obliged to deliver comprised the Electronic Point of Sale Service, the EPOSS, the Order Book Control Service, the OBCS, and the Automated Payment Service, the APS, together with the supporting infrastructure necessary to deliver those services.
The provision of those services by ICL Pathway became known as the Core Systems Release, FUJ00001524. It was planned that additional functionality would be included in a later software release, known as the Core Systems Release Plus, FUJ00001525.
The new conditions for acceptance of the Core Systems Release, CSR, were defined in schedule A11 of the Codified Agreement concluded between Post Office Counters and ICL Pathway on 28 July 1999. That’s POL00028208. In essence, this provided that Post Office Counters would be legally obliged to accept the Horizon System if, following completion of the operational live trial, known as the core observation period by now, the thresholds for acceptance of the system had been met and a timetable had been agreed between the parties to resolve any outstanding medium severity faults, known as the category B faults.
The thresholds for acceptance of the CSR, the Core Systems Release, were defined in schedule A11 at paragraph 2.2 in the following terms: the thresholds will not be met in respect of CSR acceptance if there are “one or more high severity deficiencies” known as category A faults, more than 20 medium severity category B faults, more than 10 category B faults in respect of any one CSR acceptance specification expenses.
By permitting up to 20 outstanding faults of medium severity, these new thresholds represented a relaxation of the conditions for contractual acceptance, as compared with that defined in the original authorities agreement. However, this lowering of the threshold for acceptance was somewhat mitigated by the requirement to have in place an agreed timetable for resolution of all such faults prior to acceptance of the Core Systems Release.
The timetable for the acceptance of the Core Systems Release was defined at schedule B09 of the Codified Agreement. It made provision for an eight-week core observation period to run between 31 May 1999 and the 26 July 1999, POL00028224. Upon conclusion of the core observation period, the Codified Agreement provided for a further three-week operational trial period, during which the parties would seek to reach agreement on the occurrence and severity of any faults or deficiencies observed during the operational live trial of the system: POL00028208.
The principal method by which Post Office Counters exercised its right to inspect the performance of the Horizon system during the core observation period was by raising an “Acceptance Incident”, also known as an AI. This constituted the giving of formal notice to ICL pathway of a fault or a deficiency observed by Post Office Counters during the live trial of the system. Under the terms of the Codified Agreement, ICL Pathway was responsible for assessing the severity of each acceptance incident raised by Post Office Counters, and for assigning a category to it.
Schedule A11 of the Codified Agreement made provision for resolving disputes relating to the severity assigned by ICL Pathway to each acceptance incident.
The occurrence and categorisation of acceptance incidents would ultimately be determinative of whether the threshold for acceptance of the CSR had been met.
Within Post Office Counters, responsibility for managing the acceptance process rested with an acceptance board whose core membership included John Meagher, Horizon acceptance manager, David Smith, head of automation transformation, and Bruce Mcniven, Horizon Programme Director.
Terms of reference for the acceptance board were prepared in which its duties were defined as follows: to formally advise ICL Pathway of the acceptance outcome, to submit appropriate acceptance recommendations to the managing director of Post Office Counters, for endorsement; to effect any decisions consequent on acceptance recommendations as directed; to act as an escalation platform for the resolution of disputes over the severity assessment of individual acceptance incidents; and to act as an escalation platform for acceptance issues raised at the acceptance management forum.
In order to assist the acceptance board to make an appropriate acceptance recommendation, Post Office Counters prepared a hot list, as it was called, of all medium and high severity acceptance incidents observed during the core observation period.
We can see from POL0028355 that as of 13 August 1999 the hot list comprised of three acceptance incidents assessed by Post Office to be of high severity, one acceptance incident assessed by Post Office Counters to be of medium to high severity, and 11 acceptance incidents assessed by Post Office Counters to be of medium severity.
I shall now explain in further detail the nature of the three acceptance incidents which were deemed by Post Office Counters to be of high severity.
The first of these was AI (Acceptance Incident) number 376. We can see that, no need to turn it up now, at POL00030393. This related to the integrity of the data which was fed from the EPOSS to the Transaction Information Processing. You will remember that’s Post Office’s own internal system for interfacing with Horizon in order to extract data for accounting purposes.
Post Office Counters’ analysis of the data fed into TIP revealed a discrepancy between the volume of transactions recorded by EPOSS at the counter and the branch’s weekly cash account, ie the aggregate of the weekly transactions produced by EPOSS. According to Post Office Counters the existence of this discrepancy had only come to light as a result of a temporary function performed by TIP by which it derived a cash account from the transaction data, which had extracted from EPOSS, and compared this to the cash account which had been produced by Horizon.
The Post Office assessed the business impact of this deficiency to be severe.
The second of these was AI 218, and we can see that at POL00028365. This AI related to the quality of the training course provided by ICL Pathway to assist subpostmasters in preparing their weekly cash account. It was the perception of Post Office Counters that the poor quality of ICL Pathway’s training could adversely affect the ability of end users to complete their daily balancing and produce a cash account. The knock-on effects of this deficiency, according to Post Office Counters, were that an excessive amount of time was being taken by end users, subpostmasters, to produce their weekly cash account, the volume of calls being made to Post Office Counters’ support services relating to cash account problems was unmanageable, and that significant time and resource was spent by Horizon Field Support Officers and Horizon Post Office Counters backend services in remedying user errors.
The third these incident was AI 298 which can be seen at POL00030393 and this AI related to the overall stability of the system and concerned the occurrence of lock-ups, as they were called, or system freezes, as they were called, which had been observed to occur partway through the processing of a transaction on Horizon, the only remedial action open to the end user was to reboot the system.
The deficiency was assessed by Post Office Counters as being liable to cause a substantial loss of time and business, a loss of confidence in the system on the part of subpostmasters and customers, a risk of accounting errors and a severe disruption to Post Office Counters’ operating processes.
Although ICL Pathway proposed a series of remedial measures to justify the downgrading or closure of each of these high severity AIs, the parties ultimately agreed that acceptance of the Core Systems Release had not been achieved at the conclusion of the operational trial review period. In an amendment to the Codified Agreement dated 20 August 1999, which came to be known as the First Supplemental Agreement, the parties agreed to undertake a programme of joint workshops for the purpose of agreeing resolution plans for all outstanding high and medium severity faults in the system.
Peter Copping, director of PA Consulting Group, was appointed as an expert to facilitate the effective solution of the outstanding faults and where necessary to adjudicate on any unresolved disputes between the parties.
In a timetable appended to that First Supplemental Agreement, it was planned that acceptance of the core systems requirements would be achieved by 24 September 1999.
That’s FUJ00000485.
The joint workshops which ran between 20 August and 17 September 1999 were chaired by the late Keith Baines, Head of Horizon Commercial, and Tony Oppenheim, Finance and Commercial Director of ICL Pathway.
You will be hearing from a number of witnesses who participated in these joint workshops, including John Meagher, Horizon Acceptance Manager. Mr Meagher, who was responsible for managing the joint workshops on behalf of Post Office Counters, will say:
“Pathway would invite Post Office to accept that elements of an acceptance specification had been passed or had exclusions”, ie issues that were understood and did not present a significant business impact, and therefore could be parked for later attention.
There was always disagreement on this impact assessment because Post Office always had correctly reviewed the business and operational impact, whereas Pathway, given that acceptance, was dependent on there being fewer AIs than an agreed number, tended towards lower classifications. These workshops were effective in giving Post Office leverage to force Pathway to agree to make corrections, but by 24 September 1999, not all of the outstanding high and medium severity AIs had been resolved to the satisfaction of Post Office Counters.
The most serious of these concerned the data integrity issues associated with AI 376. Pathway’s proposed remedial plan for this AI involved the development of a new piece of software known as “The Accounting Integrity Control Release”. You’ll be hearing evidence from Tony Oppenheim, Finance and Commercial Director of ICL Pathway, who will describe the release in the following terms – this is WITN03770100:
“The agreed solution [that’s the agreed solution to AI376] was a three-way reconciliation mechanism, otherwise referred to as an integrity check, that would validate that (i) the branch counters, (ii) TMS [that’s Pathway Central System] and TIP [that’s Post Office Counters Central System] were all balanced and in sync.”
Where they were not found to be in sync, the reconciliation would enable corrections to be made by either Pathway or Post Office Counters in their respective TMS or TIP systems or to bring TIP or TMS into line with the transactions logged at the counter and the cash account committed by the postmaster. Where they were found not to balance, or a difference was identified between the transactions recorded in the branch and the cash account committed by the postmaster, the intention was that the reconciliation would identify the reason and enable it to be corrected.
Given that the accounting integrity control release had not been developed and its efficacy was not yet proven, Post Office Counters agreed to accept the Horizon System in late September 1999 on the condition that additional criteria aimed at demonstrating the efficacy of ICL Pathway’s rectification plans were met. These criteria were contained in a further amendment to the Codified Agreement, which was signed on 24 September 1999, and which became known as the Second Supplemental Agreement. That’s found at FUJ00079316.
In essence, this agreement permitted Post Office Counters to suspend the rollout of the Horizon IT system in the event that ICL Pathway failed to demonstrate the effectiveness of the remedial measures, which it had agreed to adopt, in order to address Post Office Counters’ concerns.
It would appear that this decision of Post Office Counters to grant conditional acceptance of the system was not one which all of the senior managers involved in the Horizon programme shall be able to endorse.
You may wish to examine, in due course, in further detail how that decision was reached, and why it was that some senior managers believe that the effectiveness of ICL Pathway’s remedial measures ought to be proven before acceptance of the system rather than after it.
Whilst the terms of the Second Supplemental Agreement all were negotiated in late September 1989, further concerns were being voiced internally within ICL Pathway about the stability of the EPOSS application. It would appear that a report on EPOSS was produced on 21 September 1999 in which a series of recommendations were made to consider redesigning or rewriting EPOSS in order to address its “known shortcomings”.
Regrettably, Fujitsu has been unable to locate a copy of this important document. However, as a result of an internal development audit of the Core Systems Release Plus, which was conducted by ICL Pathway between 8 September and 7 October 1999, a further recommendation was made to the directors of ICL Pathway on or about 22 November 1999 to reconsider the earlier recommendations that EPOSS be redesigned or rewritten, either in whole or in part. That’s Fujitsu 00079783.
This further recommendation was made “in the light of the continued evidence of poor product quality” which had emerged from the internal audit.
The individual who assigned ownership of this recommendation was Terry Austin, ICL Pathway’s Systems Director. In his evidence to the Inquiry, Mr Austin has stated that:
“This option was debated at length by a senior manager to the ICL Pathway management and technical teams, and the outcome was that we should embark on a major exercise to target the specific areas known to be the source of most (around 80%) of the issues identified, which were error handling and printing. If this approach were unsuccessful, then a rewrite would be the only option available. However, the product did become stable and the number of outstanding defects did fall within the levels defined in the acceptance criteria.”
This explanation given by Mr Austin does not, however, appear to match the justification given at the time by Mike Coombs, programme director, who on 10 May granted authorisation to close the recommendation arising from the internal audit in the following terms:
“As discussed, this should be closed. Effectively, as a management team, we’ve accepted the ongoing cost of maintenance rather than the cost of a rewrite. Rewrites of the product will only be considered if we need to reopen the code to introduce significant changes in functionality. We will continue to monitor the code quality based on product defects as we progress through the final phases of testing and the introduction of the modified code set into live usage in the network.”
Given how important the robustness of EPOSS was to the ability of subpostmasters to account for the transactions performed over the branch counter, you, sir, will no doubt wish to scrutinise very carefully how this decision was reached by the directors of ICL Pathway, and whether or not Post Office Counters was consulted about this decision, which in principle, at least, had far reaching implications for its business.
As a result of concerns about the adverse impact that rollout of the Horizon infrastructure was likely to have upon Post Office branches over the busy Christmas period, and in the light of ongoing concerns about the integrity of the data which was being fed into by EPOSS into TIP, Post Office Counters decided to suspend the rollout of Horizon in late November 1999. This decision coincided with a growing dispute between ICL Pathway and Post Office Counters concerning the root causes of the continued cash account imbalances, which were being observed during the ongoing monitoring of ICL Pathway’s compliance with the conditions imposed on acceptance.
ICL Pathway contended that one of the dominant causes of these imbalances was inaccurate or defective reference data being supplied by Post Office Counters. We can see an example of that at POL00028561.
It’s currently unclear whether and to what extent Post Office Counters accepted that the cash account imbalances being observed between October and November 1999 were attributable to its own reference data.
In December 1999 Post Office Counters commenced negotiations with Pathway over the terms of a further amendment to the Codified Agreement, which became known as the Third Supplemental Agreement. At a joint meeting between ICL Pathway and Post Office Counters on 14 January 2000, it was noted that a decision to recommence the rollout of Horizon would not be made until further checks had been carried out by TIP to verify the efficacy of the Accounting Integrity Control Release and the terms of the Third Supplemental Agreement had been agreed.
That’s POL00028509.
The Third Supplemental Agreement was signed by both parties on 19 January 2000. That’s FUJ00118186.
According to Tony Oppenheim, the Finance and Commercial Director of ICL Pathway, this agreement tacitly acknowledged that the accounting integrity control did not provide a complete solution to the data integrity issues which had first been identified as part of AI 376.
In his evidence to the Inquiry, he will state – this is WITN03770100 – the Third Supplemental Agreement recognised that in some circumstances it would not be possible to identify what had gone wrong, and therefore what correction should be made. Such instances were to be expressly disclosed by Pathway to Post Office Counters. Although not expressly stated, the logic was that it would fall to Post Office Counters as ultimate owner of the service and of the relationship with its clients and subpostmasters, to exercise a judgment in such cases.
Sir, you will undoubtedly wish to explore whether or not Post Office Counters shared that understanding of the terms of the Third Supplemental Agreement, and if so, why it was that Post Office Counters frequently exercised its judgment in a manner which was so detrimental to the interests of its subpostmasters.
So that comes to the end of the opening on Phase 2.
Can I turn to Phase 3, please.
Sir Wyn Williams: Yes. And to save you undue stress and your voice undue stress, once we get to about 4.10 or 4.15, you choose your moment; all right?
Mr Beer: Very kind, sir.
Phase 3. Introduction. I will now turn to the issues which you will be asked to consider as part of Phase 3 of the Inquiry. They relate to the operation of the Horizon System.
And when I say “operation of the Horizon System”, that is taken to include the training, advice, and assistance provided in respect of it. It will also touch upon the awareness of bugs and errors, and the issue of how disputes about discrepancies were treated.
However, the history of these issues will also be covered in Phases 4, 5 and 6, and moreover, Phase 7 will examine the current state of the provision of training, the available advice and assistance, and the dispute resolution mechanisms that are presently in place.
This phase is divided into the following themes or issues: first, the quality of training that was provided to subpostmasters and others working with Horizon. Secondly, the advice and assistance that was made available to them, and its adequacy. Thirdly, the modifications made over the lifetime of Horizon, the reasons for those modifications, and the ways in which their success was reviewed. Fourth, the knowledge and rectification of bugs, errors and defects in the system. And fifth, the approach taken by Post Office to the resolution of disputes where discrepancies or alleged shortfalls appeared.
So addressing each of those five themes in turn. Training.
By way of introduction, the starting point to this issued should of course be our list of issues, and the relevant section is found under the Horizon IT System subheading (d), and you are asked to investigate and make findings on the following issues: what training was provided to subpostmasters, managers and assistants on the use of Horizon? Who delivered that training? Did they have adequate experience and qualifications to do so? Was the training provided to subpostmasters, managers and assistants adequate? And were there any changes or improvements to training over the years, and if so, what were they?
It is hard to overstate the importance of the issue of training to this Inquiry. Put another way, a fighter jet might be a pinnacle feat of engineering, but it is absolutely useless in the hands of someone without training, and indeed likely to be dangerous to someone who is inadequately trained.
In a similar way, if one assumed that Horizon had been rolled out in a near perfect state – and that’s an assumption – from a computing perspective, subpostmasters would still have required to be carefully trained to make the project a success.
A common theme that runs through the Horizon story is the raising of user error, ie error by the subpostmasters, as an explanation for apparent discrepancies.
Even if it was assumed that there was no fault in the design of the Horizon IT System, it would still be important to examine the cause of user error. In particular, whether the training was sufficient to minimise such errors.
A good example of this is a PEAK, which we’ll come to later, called PCO129767, to be found at POL00001264. That PEAK shows that a subpostmaster had sold some currency but had incorrectly entered into Horizon that this had been settled in cash rather than by a debit card. He sought to reverse this, but did so incorrectly, raising a discrepancy. I’m not going to delve into the issue now, but when we examine the PEAK in due course, we’ll see that it says:
“As explained, the root cause is a user error, although it is also clear that the user documentation, which is Post Office Limited’s responsibility, could be clearer.”
Here, the fact that something can be described as “user error” does not determine who is ultimately at fault for this discrepancy. You will read in the evidence numerous occasions when discrepancies were attributed to user error. Indeed, you’ve already heard a great deal of evidence on this issue as part of the human impact hearings. You’ll recall Mr Palmer, Kevin Palmer, gave evidence back on 23 February 2022. He had a career in investment banking before coming to take over his mother’s, Suzanne Palmer’s, Post Office branch after she had been suspended for alleged shortfalls.
You may recall Mr Palmer’s evidence that he did not receive training on how to balance accounts. He gave evidence on how he was often faced with discrepancies which he would spend time checking against these records. He explained that at one stage, he faced a discrepancy for over £60,000, and you may recall his evidence that he rang the helpdesk operator who said, “It was not possible for this to be as a result of system error, and simply advised me to check my figures again.”
You will wish to examine and make findings on the adequacy of the advice and assistance given, to which I will consider shortly. However, I raise this point now to highlight the adequacy of training provided to subpostmasters would be a significant and important issue, even if one assumed that the Horizon System was robust. It’s undoubtedly all the more important when the computer system is not itself is not robust.
I’ve set out the issues that the Inquiry is asked to consider in respect of training, and why they are important.
Can I now deal with the following issues. Firstly, briefly cover the scale of the training project that was to be undertaken on rollout. Secondly, introducing the training programme that was designed to be delivered across the Post Office Network during the national rollout, which training was provided by ICL Pathway. At this time, I’ll introduce some of the themes and particular issues that you may wish to consider in respect of training generally. Then thirdly, I’ll summarise briefly the timeline of the training after rollout which was provided by Post Office, not ICL Pathway.
So the scale of the training project.
The rollout of Horizon brought with it the formidable task of training the entire Post Office Network to use the system. Two points merit specific mention at this stage. The first is the nature of the system that trainees were being asked to use. The task must be viewed in context, and we must cast our minds back to how subpostmasters operated in the late 1990s. You will hear helpful evidence from the Inquiry’s expert, Charles Cipione, on this issue. He will paint a picture of the state of technology generally at this time.
He notes that in 1998 it was estimated that only 31-33 per cent of UK adults had a personal computer at home. Indeed, the Post Office itself carry out a survey of 100 subpostmasters, counter-clerks and assistants in 1999, which showed that 65 per cent of them had used a computer before, and 83 per cent of them were “familiar with a keyboard”.
These are numbers that would be unheard of today, and so it’s not uncontroversial to say that the Horizon System was revolutionary for the time, and all the more so for the work of the subpostmasters.
The second issue to consider is the widespread changes that were being made to the day-to-day processes of branches up and down the country by the mere introduction of an automated system.
Horizon introduced a mandatory centralised system of work that subpostmasters would have to follow precisely in order to operate their branches. Once rolled out, each subpostmaster would have to use the same systems and processes to sell a stamp or to settle a cash account.
You will hear evidence on how much a shift this was for subpostmasters. In particular, you’ll read about a perceived lack of “conformance culture” across the Post Office Network prior to the rollout of Horizon.
I would suggest that these changes in business practice merit careful consideration by the Inquiry, along with the more obvious changes bought about by the introduction of the computers themselves. Therefore, in order to investigate the adequacy of the training provided to subpostmasters, it will be necessary to consider how the parties assess the baseline competences and training needs of the Post Office Network. It will be necessary to ask whether the training programme implemented was sufficient to meet those needs.
In terms of training on rollout and contractual responsibilities, the Post Office contracted the responsibility to train the Network to use Horizon to ICL Pathway. It will be necessary to consider the effect that this had on the adequacy of the training programme delivered. One can see that ICL Pathway would likely have had a better working knowledge of the Horizon System in order to design the training programme. It will be necessary, however, to investigate the extent to which the training took account of Post Office’s businesses, the experiences of its network, and its day-to-day operations, particularly in respect of the alleged lack of conformance culture.
It’s helpful at this stage to note that ICL Pathway subcontracted aspects of its training responsibilities to a related company called Peritas Limited, Peritas changed its name during the period, so you may also see it referred to in the documents as “ICL Training Services Ltd” or “Knowledge Pool Ltd”.
Can I turn, then, to examine some of the contractual responsibilities for training on rollout, and can we please bring up POL00028212 and can we turn, please, to page 24. Perfect.
This, just to explain, is part of Schedule A15. You can see it at the top of the codified agreement between ICL Pathway and Post Office, and this part of the schedule sets out Post Office’s requirements in respect of the delivery of Horizon.
If we look at Requirement 531 at paragraph 1.65.1, ICL Pathway agreed to deliver a training programme that would “enable Post Office’s target audience to achieve the acceptable standards in key competences as defined by Post Office Counters Limited”, with competence to be “measured by a method agreed between Post Office and ICL Pathway on a date consistent with the project plan”.
Then over the page, please. Can we look at Requirement 534, thank you, at paragraph 1.68.1.
This requirement, Requirement 554, provided that ICL Pathway was to agree the training requirements for different groups such as counter managers within the Post Office.
It may be that the terms “acceptable standards” within this clause and “competence” leave much remaining open in respect of what level of training was actually required. But you will hear evidence in due course on how those concepts of what are acceptable standards and what competence meant were fleshed out, and how that impacted the training to be delivered.
The agreement, this agreement, went on to give further detail on ICL Pathway’s obligations in Requirement 915, but I’m going to turn to that provision in more detail later.
The training provided on rollout. We can take that document down. I’ve spoken so far about the scale of the training project and the division of labour as between ICL Pathway and Post Office. You will hear evidence in due course on what training was actually provided to subpostmasters. Can we turn, please, to FUJ00001357. Page 9, please.
That might actually be page 10. Page 10 of the pdf, page 9 of the document. Thank you.
This document, as a whole, is the 11th version of the Training and User Awareness Baseline document. This is a contract controlled document setting out ICL Pathway’s response to meeting Post Office’s training needs. If you can scroll down to the diagram, this sets out the structure of the training programme. You can see that times in days or hours are allocated to parts of the training programme that were to be delivered as part of the national rollout.
And down the right-hand side, a timescale by reference to Day Zero, which is at the bottom, ie the date by which they must be provided. So subpostmasters would be invited to training events as part of the rollout of Horizon at their branch. These training events would sit alongside the branch preparations to the conversion to Horizon itself.
If we take that off the screen, the first formal training event was a non-mandatory user awareness training event. That was described as a two-and-a-half hour lecture style event that would provide an overview of the function of Horizon itself and the training programme. The second formal training event was described as “user training”, and that training had to take place within five days of Horizon going live at that person’s branch. That was subsequently elongated to up to ten days before.
Either way, the core training for the subpostmaster took place, you’ll note, within a relatively short period of time before Horizon going live at their branch. That user training is described as “a classroom-based event” where users would be instructed on how to use Horizon.
You’ll hear evidence that subpostmasters were offered the course for managers, which included training on the accounting elements for Horizon, such as balancing, and that that training lasted for one and a half days.
The counter assistants were offered a day-long course which did not cover balancing in detail. I’m not going to go into the detail of the content of the training programmes at this stage for two reasons: firstly, it’s clear that the training changed over the course of time, and it wouldn’t be a helpful use of time to chart the changes in the course of this opening. They will be matters for you to consider in the Phase 3 hearings. And secondly, there will, I anticipate, be disagreement between witnesses over the facts during this period of rollout and indeed thereafter, and it will be important to read and hear those witnesses in full.
Instead, in opening, can I just briefly identify the number of themes and issues which you may wish to grapple with when investigating training as part of Phase 3.
Firstly, the assessment of baseline competences and the analysis of training needs. Secondly, the adequacy of time made available for training and its content. Thirdly, collection and use of feedback. And fourth, the assessment of competence.
Sir, I notice the time. Having teased you with those four topics, including the assessment of baseline competences –
Sir Wyn Williams: I think that’s to think about tonight, Mr Beer.
Mr Beer: I’m afraid we’ll have to come back to it tomorrow.
Sir Wyn Williams: Fine. What time are we starting tomorrow?
Mr Beer: Can we start at 10.00 am, please, sir?
Sir Wyn Williams: Certainly. As far as the future programme is concerned, in inverted commas, we had a diversion of about an hour and a half this morning, but that won’t affect or shouldn’t affect the progress of opening this week?
Mr Beer: We always had a half day spare at the end on Friday afternoon, and so we should complete all openings by the end of Friday.
Sir Wyn Williams: Excellent. I think everybody now knows that for personal reasons I can’t be present in the Inquiry room every day, but of course I’ll be remote when I’m not in the Inquiry room. But just in case there are people who will suddenly see a blank space and wonder what’s happened, I’m still with you in spirit, if not in body.
Mr Beer: Thank you, sir.
(4.13 pm)
(The hearing adjourned until 10.00 am the following day)