In the story so far, we have seen various red flags that might have prompted POL to pause for thought about Horizon and potential miscarriages of justice and legal difficulties. These include Jason Coyne’s report in the Cleveleys case; Anne Chambers raising causes for concern in an ‘afterthoughts’ document she wrote after giving evidence in the Castleton case;Footnote 1 and the RPM note in the Misra case. In 2009, Rebecca Thompson published a story for Computer Weekly critical of Horizon. (Sir) Alan Bates created the Justice for Subpostmasters Alliance (JFSA) in 2009 and political pressure built up, with James (now Lord) Arbuthnot MP leading the charge. In 2011, Nick Wallis reported for BBC’s Inside Out South on Jo Hamilton’s and Seema Misra’s stories (while the latter was still behind bars), and Private Eye picked up the story for the first time.
In 2010, POL had considered an independent review of Horizon, but Rob Wilson, head of POL’s criminal lawyers, reacted adversely to the suggestion: ‘An investigation will be disclosable as undermining evidence,’ he wrote. Horizon would be seen as ‘a compromised system with untold damage to the business’, and ‘the reputation of my team’ might be compromised if they were thought to be pursuing cases improperly.Footnote 2
Instead, Rod Ismay, POL’s Head of Product and Branch Accounting, produced an internal report on Horizon. Largely, it was a collation of evidence and arguments in support of Horizon. He agreed it presented ‘one side of the coin’ when cross-examined about it at the inquiry, but rejected the Chairman to the Inquiry’s suggestion it was a whitewash.Footnote 3 The report blamed shortfalls on theft and cited successful prosecutions as proof of Horizon’s reliability.Footnote 4
The Ismay report reiterated concerns about carrying out a more independent investigation: it would be disclosable in court, potentially halting prosecutions and prompting appeals, a view attributed to Wilson, whose stance was perhaps best summed up in this exchange during his inquiry testimony:Footnote 5
‘Mr Beer: Was it your view that the Post Office should just grit its teeth and get on with prosecuting people?
Mr Wilson: ‘I think so, yes.’
Second Sight
Despite this resistance, a more independent review was conducted. POL acquired a new Chair, Alice Perkins, in July 2011, and the (then) MP James Arbuthnot approached her in early 2012 about problems with Horizon.Footnote 6 Paula Vennels was also a new CEO, having already held a senior role in the business for several years. Consideration was given to asking one of the Big Four accountants, Deloitte, to conduct a system-wide review but POL opted instead for case-by-case investigations that would focus on individual complaints conducted by forensic accountants Second Sight.
It was Second Sight that first uncovered prosecution flaws, including reports showing no evidence of dishonesty in cases like Jo Hamilton’s. And it was Second Sight that learned from Gareth Jenkins that Horizon had known bugs that had not been disclosed. It was this discovery that led to the Clarke Advice, as we will see.
Second Sight produced an interim report in July 2013. Its preliminary conclusions ‘found no evidence of system wide (systemic) problems with the Horizon software’, but also ‘2 incidents where defects or “bugs” in the Horizon software gave rise to 76 branches being affected by incorrect balances or transactions’. They found that ‘POL’s response [to SPM problems] can appear to be unhelpful, unsympathetic or simply fail to solve the underlying problem’, and noted a ‘lack of an effective “outreach” investigations function within POL that results in POL failing to identify the root cause of problems and missing opportunities for process improvements’, as well as other problems.
POL amplified the claim about not finding systemic defects in the software into a positive claim of Horizon reliability, downplayed the significant problems surrounding the management of the system, and sought to undermine Second Sight. But by the time of the inquiry, POL admitted that it was defensive and closed-minded about Horizon’s faults, had constrained Second Sight’s remit, and had withheld key information. While accepting some responsibility, POL also blamed external voices, including reproach for Andy Parsons and another solicitors’ firm, Linklaters, as well as Fujitsu, for also criticising Second Sight.
After the interim report, the evidence suggests POL shifted from allowing Second Sight access to legal files to obstructing them. Second Sight stated that they were routinely denied access to the files as they were considered legally privileged. POL lawyers also referred them to the non-disclosure and confidentiality requirements outlined in their contract to inhibit them in their work.Footnote 7 The time of this shift coincided with the departure of Susan Crichton, the then General Counsel, from POL, and her replacement with Chris Aujard as Acting General Counsel.
The Independent Mediation Scheme
At around this time, POL created a mediation scheme. The scheme appears partly to have been an attempt to manage Second Sight by making them operate within the context of a working group, on which POL sat, and creating a mechanism for resolving cases without a definitive investigation. POL’s expectation was that most cases would be resolved without any compensation paid or with minimal token payments (following advice from the solicitors’ firm Linklaters).Footnote 8 In their final submissions to the inquiry, they recognised that they were likely to be criticised for persisting with a lengthy and costly scheme ‘outwardly stated’ to be about resolving complaints that prioritised complaint resolution ‘over truth seeking’. They claimed a ‘not insignificant portion’ of cases settled, though the actual figures – 24 out of 136 cases, just short of one in five – were mediated and we do not know how many of them gave rise to agreements within which SPMs were properly compensated.Footnote 9 They concede the scheme was run in a way inconsistent with their original commitments to Sir Alan Bates and the JFSA.Footnote 10 SPMs would say the complaints mediation scheme was run with the hallmarks of obstruction and attrition.Footnote 11
Importantly, Sir Anthony Hooper, who chaired the scheme’s working group, says he told POL that it was Horizon, not the SPMs, that was probably the problem:Footnote 12
I tried to make it clear to Paula Vennells and to the Chairman that the Post Office case didn’t make sense, and I felt that throughout, and no doubt Second Sight did. It didn’t make sense that reputable SPMs appointed by the Post Office, after an examination of their characters, would be stealing these sums of money.
It didn’t make sense, in particular because within a matter of days of any alleged theft, they had to balance the books. It just never made sense. I made that point over and over again.
POL kept that quiet for years. It seems likely that Hooper was subject to a confidentiality agreement until POL agreed to lift the obligation on him, which enabled him to speak to the journalist Nick Wallis in 2021.Footnote 13
Arguments persist as to why the scheme was ended, but one issue was POL’s eventual decision to exclude all criminal cases from the mediation. Lord Arbuthnot, who had helped broker the Second Sight investigation, regarded this as antithetical to the scheme, which had been set up to help SPMs with convictions, such as his constituent, Jo Hamilton.Footnote 14 POL said they had formed the view that the criminal convictions were safe and there was no evidence Horizon was at fault for losses and so convicted claimants should be excluded.Footnote 15 The basis of the judgment that convictions were safe lies in a process that remained in the shadows until the Hamilton case in 2021. We now turn to that work on prosecution safety.
Reviews Relevant to Prosecution Safety
Simon Clarke, a barrister with Cartwright King solicitors in 2013, told the inquiry that towards the end of June 2013, either Jarnail Singh or Rodric Williams contacted his colleague, Martin Smith, who told Clarke that ‘Second Sight are going to say there are two bugs’. His initial reaction was ‘Who told them?’Footnote 16 Suspecting Gareth Jenkins, he and Martin Smith contacted Jenkins about a live case with which Jenkins was involved to determine if they were correct. Interestingly, they covertly recorded the conversation. Jenkins informed them about the bugs and mentioned that he had discussed them with Second Sight. Clarke writes a written note on the live case on 1 July saying (amongst other things) ‘in any event I require a face-to-face conference with Gareth Jenkins upon publication of the Second Sight report’.Footnote 17 That conference never took place.
Clarke needed to assess what advice to give for POL to meet its obligations as a prosecutor under R (Nunn) v Chief Constable of Suffolk Constabulary [2015] AC 225: that case says ‘There can be no doubt that if the police or prosecution come into possession, after the appellate process is exhausted of something new which might afford arguable grounds for contending that the conviction was unsafe, it is their duty to disclose it to the convicted defendant.’ Rather than meeting Jenkins and determining the extent and cause of the problems, Clarke decided to review Jenkins’ witness statements on past cases and compare them with what he had recently said about bugs in the system. He then wrote the July advice criticising Jenkins’s failure to disclose such bugs in his evidence.
Mr Clarke could not explain to the inquiry why he did not speak to Jenkins as he had first indicated he thought he should, saying he had probably decided by then ‘to look forwards rather than back’.Footnote 18 The last two sentences of his July advice on Jenkins, given around the time he had decided to look forward and not back, read:Footnote 19 ‘In an appropriate case the Court of Appeal will consider whether or not any conviction is unsafe. In so doing, they may well inquire into the reasons for Dr Jenkins’ failure to refer to the existence of bugs in his expert witness statements and evidence.’ A vulnerability that he knows the Court of Appeal might want to explore lay unexamined.
The inquiry also probed another possible explanation for not asking Jenkins what had gone wrong. Cartwright King had prosecuted a good many cases for POL from 2012 that had not involved Mr Clarke personally, including cases involving evidence from Gareth Jenkins. The evidence suggests that like Warwick Tatford, they too handled him inadequately, and perhaps on occasion improperly.Footnote 20 There was evidence before the inquiry that Martin Smith and Rodric Williams may have discussed on 2 September 2013 that they ‘don’t think he’s ever been advised of his duties’.Footnote 21 Cartwright King lawyers claimed either not to have known about the problem with instructing Jenkins or not to have appreciated that there was a conflict of interest.
Clarke’s advice on Jenkins led to his firm conducting a review of all prosecutions from 2010 onwards. It was confined to cases from 2010 onwards on the basis only Horizon Online had been identified as producing undisclosed errors and considered the need to disclose two documents only. One was Second Sight’s interim report and the second was a report from Helen Rose. That Gareth Jenkins’ evidence had been found to be tainted was not considered for disclosure.Footnote 22
POL suggests the inquiry evidence shows that Cartwright King was a poor choice to conduct both prosecution work and the review work. One of their lawyers had ‘accepted that by 2013 [Cartwright King] was “not covered in glory, as far as their role as independent lawyers [was] concerned”’, key staff ‘had limited prosecution experience, which, “plainly wasn’t a safe state of things”’, and their approach was ‘somewhat disjointed’ in ways ‘plainly unsatisfactory’.Footnote 23 They say too that ‘POL was an “enormous” client for Cartwright King and that this factor “must have” played a part (whether consciously or unconsciously) into how Cartwright King approached its advice and its ability to advise impartially and independently.’Footnote 24 POL concede that it is ‘very clear’ Cartwright King should not have been instructed, ‘given [Cartwright King]’s lack of relevant expertise and management of the POL account, and the devastating consequences of such’.Footnote 25 And they accept that they failed to supervise and monitor their work properly, pointing the finger at Jarnail Singh, POL’s in-house criminal lawyer at that time.Footnote 26
They might have thought about pointing the finger elsewhere. When the Cartwright King review was first suggested, POL General Counsel (GC) Susan Crichton said she ‘feels as if there is a conflict here which I am not sure I understand’.Footnote 27 The problem was Cartwright King was being asked to mark their own homework. That led to the involvement of Brian Altman KC.
The Altman Review
The response to Crichton’s concern about a conflict was to instruct Brian Altman KC to review Cartwright King’s review. What might properly be described as a solicitor–own client conflict that prevented them from acting, Altman described in less serious terms as ‘the possibility of the suggestion of a commercial conflict of interest. Given CK’s professional’ approach to the review, he rejected ‘any suggestion’ that were not or could not act with ‘an independent and professional approach to the Horizon issues’.Footnote 28
This is one of many potential problems with the Altman review.Footnote 29 Two points merit particular mention. Similar to Mr Clarke, Mr Altman began his work by thinking he might need to talk to Jenkins to avoid criticism of his report but persuaded himself he did not. He denies the suggestion put to him by Jason Beer KC that this was ‘loading the dice by excising him in this way’,Footnote 30 but accepts that he should have looked at the process by which Jenkins was instructed ‘with the clarity of hindsight’.Footnote 31 He also accepts he should have disclosed his and Simon Clarke’s assessment that Jenkins was a tainted witness.Footnote 32 Altman tells the inquiry he would have done things differently if it had been a CPS case; his failures were simply a mistake:Footnote 33
I’m human, like the rest of us, Mr Stein. Probably including you and I make mistakes and we all make mistakes and, as I said earlier, I would like to say was misjudgement. I don’t think I made the judgement about it, it passed me by, I don’t know why I didn’t see it. It’s regrettable and I regret that I didn’t.
Submissions from one of the SPM legal teams, led by Ed Henry KC, suggested:Footnote 34
Neither Mr Altman KC nor Mr Clarke has offered a compelling or even clear explanation for their mutual failure to provide the Criminal Cases Review Commission (CCRC) and Mrs Misra with disclosure of the Gareth Jenkins’ ‘taint’. We submit that the explanation for this is simple but alarming: the Altman General Review and the Cartwright King sift review were ‘optic pieces’ (akin to Hugh Flemington’s strategy ‘to give off the signals that we are proactive, doing all the right things’). They were designed to make it appear as if the Post Office was complying with its duties of disclosure, in order to keep the CCRC and claims for compensation at bay. It is therefore unsurprising that their sifts and reviews did not lead to the Post Office swiftly restoring matters to the Court, and providing the CCRC with a clear and comprehensive appraisal of Gareth Jenkins’ misleading evidence, and the Post Office’s abject failure to instruct him.
In their submissions to the inquiry, POL acknowledged that limiting the review to 2010 became ‘difficult to sustain’ and should have been recognised as such by Cartwright King.Footnote 35 Patterns in the cases being reviewed by Cartwright King lawyers ‘ought to have put them on notice’ of significant problems.Footnote 36 Furthermore, their own guidance was flawed, and there was ‘significant inconsistency’ in advice on cases, which was sometimes wrong.Footnote 37
Despite these problems, Brian Altman’s General Review concluded in 2013 that Cartwright King’s sift review was ‘fundamentally sound’.Footnote 38
Throughout the life of this story, Altman advised on many occasions. On 2 August 2013, he provided an ‘Interim Review of Cartwright King’s Current Processes’. Beer KC tells the inquiry in his opening that ‘Mr Altman considered two copy prosecution files, something that does not appear later to have been reflected in a submission to the Court of Appeal in Hamilton, namely that this advice considered a review of the process, though not the individual decisions in reviewed cases.’Footnote 39
He advised on risks in the mediation (9 September 2013), warning that ‘lawyers acting for those individuals may be using the scheme to obtain information which they would not normally be entitled to to pursue an appeal.’ His General Review was dated 15 October 2013, and on 19 December, in a review of the PO’s role as a prosecutor, he concluded that there was no good reason to recommend that the PO should discontinue its prosecution role. Counsel to the inquiry suggests POL at this stage was leaning significantly on Altman’s reputation as a former ‘First Treasury Counsel’, and in March 2015 Mr Altman gave ‘important advice’ on the charging of theft and false accounting. That sought to ensure a questionable statement made to Second Sight that theft and false accounting were equivalent charges was ‘defensible’ when it (arguably) was not. Second Sight was concerned that theft charges were being used to leverage false accounting pleas. Had the two charges been equivalent, their concern would make no sense. Mr Altman advised in July 2016, having reviewed eight cases, that ‘the allegation that [the] Post Office had operated a deliberate policy to charge theft, when there was no or no sufficient evidence to support it, just to encourage or influence pleas of guilty to charges of false accounting is fundamentally misplaced’.
Mr Altman also provided an ‘Advice on settlement’, dated 17 July 2019, advising against ‘any public apology that risks misinterpretation or the implication of an admission of fault’ because it might lead to a greater risk that convictions would be seen as unsafe. Advice on the Bates judgments noted that ‘the apparent late concession’ on remote access was a ‘startling concession’ and requested further information in that regard. Evidence before the inquiry suggests other incidences of perhaps less formal advice. POL notes one example in its final submission to the inquiry: a letter sent to the CCRC and reviewed by Altman, which he accepted had omitted ‘important information’.Footnote 40
Brian Altman represented the PO before the Court of Appeal in 2020 and was part of the team that resisted the arguments that POL’s conduct of prosecutions was an affront to justice.Footnote 41 Without suggesting this was the intention, there is some indication that, had the team succeeded, the Clarke Advices on Gareth Jenkins and shredding may not have become public knowledge. One could speculate, too, on whether the inquiry would have been granted statutory powers.
Filtering Out and Up
Evidence before the inquiry suggests various ways in which detailed, troubling information was sanitised at source or when it was passed up the corporate chain. Early in Cartwright King’s review, allegations of shredding disclosure records surfaced. Simon Clarke advised this could amount to perverting the course of justice.Footnote 42 It was then referred to only crypitcally as ‘cultural issues’ in POL documents, and there is no evidence of any substantial investigation into what happened.Footnote 43 Altman’s review was used to advocate to the CCRC that their post-conviction disclosure was being properly handled.Footnote 44 Paula Vennells wrote to a government minister in 2015 to say that ‘Through our own work, and that of Second Sight, we have found nothing to suggest that, in criminal cases, any conviction is unsafe.’Footnote 45 She had similarly defended the propriety of POL prosecutions before a Parliamentary Select Committee.Footnote 46
Neither the Second Sight interim report nor the Jenkins Problems were properly canvassed with the board. They did not see Clarke’s Jenkins advice or a summary of it. They had Second Sight’s interim report and a cover paper on it from Susan Crichton, but failed to properly scrutinise it, focusing instead on what positive news there was in the report and on concerns about the management of the review by Crichton.Footnote 47 POL Chair Alice Perkins appeared to be concerned that Susan Crichton had not sufficiently ‘marked’ Second Sight when they produced their interim report.Footnote 48 Both POL’s Chair and CEO (Vennells) criticised Crichton, and she was humiliatingly excluded from the board meeting and left POL not long afterwards. Her exclusion prevented her from discussing matters further. Her paper ‘did not highlight the critical findings which needed to be discussed’.Footnote 49 However, it has been suggested that the board might (and could) have put two and two together from a significant litigation report that Susan Crichton also prepared for it.Footnote 50 Under her watch, the POL legal team failed to take adequate steps to investigate the shredding allegations and apparently either minimised or failed to draw it to the attention of Paula Vennells; POL submitted to the inquiry that this might represent a ‘serious downplaying of the issue’.Footnote 51 Crichton relayed incorrect information about an internal audit having been reviewed by Deloitte and sought to give it privileged status when it should not have had it.Footnote 52 She was also noted to have said that all prosecutions had been successful, a false narrative likely derived from the Ismay report or a briefing from one of her criminal law colleagues.Footnote 53
A summary of a report from Deloitte (the Project Zebra Report of May 2014) disclosed critical information on the vexed issue of remote access. The Board were not told. Acting GC Chris Aujard said Roderick Williams had carriage of the issue but accepted that the board report which he had co-authored was too brief and should have included information on remote access.Footnote 54 He blamed pressure from Paula Vennells and his workload. Failure to disclose this ‘highly relevant information’ was not deliberate, POL say, but arose from a failure to understand the report.Footnote 55
After Vennells’ uncomfortable appearance before the Select Committee in 2015, incoming POL Chairman Tim Parker was encouraged to get on top of any problems with POL prosecutions by the PO Minister. It was decided to instruct Jonathan Swift QC (now a judge) to conduct another review. His report had problemsFootnote 56 but put its finger on many of the key issues, including the POL having publicly stated incorrect positions on remote access. It may have been a not so subtly coded way of wondering if POL had been lying.
The full report was kept from the board and the minister, a move which was at least partly influenced by advice on the risks of losing legal privilege, advice that could be criticised as wrong or incomplete.Footnote 57 A letter reporting the summarising of the Swift Review to the Post Office Minister, drafted with the help of Jane Macleod (GC until the Bates litigation went pear-shaped), similarly failed to include information from Swift’s review on this that POL had ‘failed to give the full picture as to Horizon’s remote access’.Footnote 58 Swift himself, though, had reviewed a draft. Macleod signed the Bates defence, which stated an incorrect position on remote access as GC on its submission to the court, saying she had taken comfort that the ‘fact heavy section of the defence … had been appropriately reviewed by those with relevant expertise’.Footnote 59
Assertions of Horizon’s reliability and the propriety of convictions were not confined to the courts or Parliament. Notably, journalists have reported being threatened with costly libel proceedings for suggesting problems.
The ‘do not investigate’ approach ran deep. In 2012, Richard Morgan advised at a conference, where, according to a note from that conference, advice was given against POL commissioning an independent report and suggesting a PR-focused approach. Morgan accepts he might well have advised against conducting an investigation (whilst saying this was a preliminary discussion, and not considered advice).
It is not suggested that the PR strategy was his idea. Yet, in that note we see the kernel of POL’s strategy during the Second Sight investigation and subsequent mediation.
Whatever the findings of the expert report it will not resolve the problem. POL will be ‘damned if they do and damned if they don’t’. If the findings are that there are no issues with Horizon people will see that as a ‘whitewash’ whereas if the findings are negative that will open the floodgates to damages claims by SPM’s who were imprisoned for false accounting and Access Legal will start to pursue all the civil cases they are currently sitting on.
POL will always have this problem – some people will never trust computers and will always believe they have an inherent problem.
A less risky approach is to agree to take the relevant MP’s privately through particular cases in which they are interested.
POL needs to engage with its stakeholders by perhaps sending out a questionnaire about Horizon to SPM’s getting their views and seeking to address the more sensible ones. This is more a PR exercise.
Between this conference in 2012 and the Bates litigation that we began this account with in Chapter 3 lay the Second Sight review, the Cartwright King review, the Altman and Swift reviews, and the abandoned mediation scheme. No less a figure than a retired Court of Appeal judge had told POL that their idea that SPMs were stealing money that was highly likely to be picked up in the weekly balancing of accounts made little sense, and that it was much more likely that Horizon was to blame. Still POL and their cohort of helpful lawyers ploughed on. We turn now to consider how lawyers help their clients to see the world in such flawed ways.
Open access