England and Wales High Court (Administrative Court) Decisions

IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
ADMINISTRATIVE COURT

		Royal Courts of Justice Strand, London, WC2A 2LL
		16/05/2013

Mr Justice Nicol :

1. On 19^th November 2010 there was a meeting between representatives of Her Majesty's Revenue and Customs ("HMRC") on the one hand and Goldman Sachs International and Goldman Sachs Services Ltd (except where it is necessary to distinguish between them, I shall refer to them collectively as "Goldman Sachs") on the other. There were a number of outstanding and long running disputes between HMRC and Goldman Sachs. Among these was an issue as to whether Goldman Sachs was required to pay National Insurance Contributions ("NICs") in consequence of an arrangement they had with some of their employees. A related question was whether, if NICs were payable, Goldman Sachs should also pay interest on the outstanding contributions. The arrangement which Goldman Sachs had entered into with their employees was one which a number of other companies had also adopted. HMRC had engaged in a similar dispute with them. However, all but Goldman Sachs had settled with HMRC in 2005 on terms that they would pay 100% of the claimed NICs, but no interest. Goldman Sachs continued to contest liability, in part on the basis that HMRC had assessed the wrong company within the group.
2. At the meeting on 19^th November 2010, HMRC was represented by, among others, Mr David Hartnett who was, at the time, Permanent Secretary for Tax at HMRC and one of the Commissioners. It was his understanding that the outcome of the meeting was that there had been agreement on all issues. In particular, Goldman Sachs had promised to pay 100% of the NICs; HMRC had promised to forego any interest on the NICs. Some of the remaining issues were settled in favour of HMRC, others in favour of Goldman Sachs.
3. HMRC accepts that their representatives made mistakes at this meeting. First, Mr Hartnett was under the impression that there was a barrier or potential barrier to HMRC recovering interest on the unpaid NICs. He had not consulted the lawyers litigating the matter on behalf of HMRC about this. His recollection was flawed. Second, internal HMRC procedures required settlements of over £100 million (and where the Department was proposing to concede one or more of the issues) to be approved by the High Risk CorporatesProgramme Board ("the Programme Board"). That was not mentioned at the meeting of 19^th November 2010. Mr Hartnett and the other representatives of HMRC had not appreciated this would be necessary. They should have done. As I have said, Mr Hartnett believed an agreement had been concluded on 19^th November 2010.
4. When the Programme Board met on 30^th November 2010 it approved (retrospectively) all elements of the agreement with Goldman Sachs except the concession to forego interest on the NICs. When Goldman Sachs was told about this, they were agitated. They said that they believed that they had a concluded agreement with HMRC from which it could not or should not resile.
5. The arrangements within HMRC allow the Commissioners to act through two of their number. As I have said, Mr Hartnett was one of the Commissioners. On 9^th December 2010 he met with another, Melanie Dawes, who was also the Director General for Business Tax within HMRC. In addition, Freda Chaloner, the chair of the Programme Board, was present and took part in their discussion, but it was Mr Hartnett and Ms Dawes who decided to approve and endorse the 19^th November settlement agreement with Goldman Sachs.
6. Some of these matters were disclosed by a lawyer working within HMRC in the course of 2011. That led to hearings by the House of Commons Public Accounts Committee and the Committee's 61^st Report HC 1531 on 14^th December 2011.
7. The Claim Form in these proceedings was issued shortly afterwards on 22^nd December 2011. The Claimant campaigns against what it sees as the harmful effects of tax avoidance and the use of tax havens. HMRC are the Defendants. The two Goldman Sachs companies are named as Interested Parties, although they have taken no part in the proceedings. The Claimant challenges the decisions to enter into the agreement on 19^th November 2010 and to ratify the agreement on 9^th December 2010. Originally, the Claimant sought a quashing order in respect of those decisions. However, Simon J. was prepared to grant permission only for the Claimant to seek declaratory relief. He subsequently granted a protective costs order with a similar limitation. The Claimant has not sought to enlarge the nature of relief beyond a declaration.
8. Subsequent to the issue of these proceedings, the National Audit Office ("the NAO"), which is led by the Comptroller and Auditor General, conducted a review of the settlement of large tax disputes. By the Exchequer and Audit Departments Act 1921 s.2 the Comptroller and Auditor General is required to examine the accounts of HMRC to ascertain that adequate regulations and procedures have been framed to secure an effective check on the assessment, collection and allocation of revenue and that they are being duly carried out. The report examined 5 cases in particular. It anonymised them, but Mr Eadie QC, on behalf of the Defendants, agreed that one of these cases, Case E, was the Goldman Sachs settlement. The NAO commissioned a retired High Court Chancery Judge, Sir Andrew Park, to provide advice. This led to a report by the Comptroller and Auditor General, Amyas Morse, on 14^th June 2012 (HC 188). It will be necessary for me to say more about this report in due course.

The Litigation and Settlement Strategy grounds of challenge

9. The Claimant relies on the HMRC policy which is published under the title "Litigation and Settlement Strategy". The version which was in force in 2010 had been published in 2007 and it is that which is relevant to the present matter. It was later revised in 2011. Paragraph 1 of the 2007 edition says, "This guidance sets out principles for bringing tax disputes to a conclusion whether by agreement with the taxpayer, or by litigation." It continues:

"Settlement by Agreement

13. Settling disputes by agreement allows both HMRC and our customers to avoid the expense and delay of litigation and so agreed terms for settlement should always be sought before going to litigation. However, settlement terms must be consistent with the reasons for undertaking an enquiry in the first place, which are to influence taxpayer behaviour positively and to challenge behaviours that contribute to the tax gap [the tax gap is the gap between tax which is due and that which is paid]. The guidance below indicates what represent sufficient settlement terms.

Settlement terms

14. General Principles

Deal with each dispute on its own merits. Do not enter into "package deals", in which a range of issues are settled for a single payment that is not subdivided amongst individual disputes.

Some disputes have an all-or-nothing character, involving a single point of law that would be decided one way or the other by the courts, with no middle ground. Such disputes should be settled on all-or-nothing terms: do not split the difference or offer any discount for an agreement not to litigate.

….

Do not undercharge tax, interest or penalties in the interest of quick settlement, even if doing so would provide a good return on time spent on the case. Always consider whether settlement terms do enough to promote positive customer behaviour and deter non-compliance.

15. In avoidance cases

….

If our advice is strong, do not accept settlements for less than 100% of the tax and interest due." [emphasis in the original]

10. The Litigation and Settlement Strategy was published on the HMRC's website and its importance was promoted by, among others, Mr Hartnett. In the magazine Tax Journal for 11^th June 2007, he wrote that, for the most part the policy aimed to articulate principles which should underpin what would inevitably be questions of judgment, but it also contained "some rules which should be followed strictly…bright lines, not to be crossed." They included, "no horsetrading between unrelated points and no package deals that settle a range of issues for a single undifferentiated sum of money and, where there was an all-or-nothing dispute, there should be an all-or-nothing settlement."
11. The Claimant argues that, because the Litigation and Settlement Strategy was a published policy, HMRC was obliged to follow it in order to ensure fairness between taxpayers. Furthermore, it was for the Court to decide what the policy meant - see for instance R (Raissi) v SSHD [2008] QB 836 [120] – [123] and Tesco Stores Ltd v Dundee City Council [2012] UKSC 13 at [18]-[20].
12. The Claimant's case is that the agreement on 19^th November 2010 infringed this guidance. Contrary to paragraph 14 it was a package deal which traded a promise to pay 100% of the NICs for HMRC's promise to forego interest on those contributions. Principal and interest were effectively a single issue. In county court proceedings against Goldman Sachs which had been issued in 2003 the Revenue claimed both. The 19^th November agreement "split the difference", contrary to paragraph 14. Likewise, contrary to paragraph 15, this was a situation where HMRC's case was strong, but it had accepted a settlement for less than 100% of the tax and interest. Furthermore, Goldman Sachs had gained an advantage over the companies who settled with HMRC in 2005. It had retained the money which was due to the Revenue for another 5 years without having to pay interest. It had done so because of its aggressive behaviour. This settlement did the opposite of encouraging taxpayers to behave positively and was therefore contrary to paragraph 13 of the Litigation and Settlement Strategy.
13. In choosing to ratify the November agreement on 9^th December 2010, the Claimant submits that HMRC failed to take account of the Litigation and Settlement Strategy which was a relevant consideration and this decision was therefore also tainted by a failure to take account of a relevant consideration.

Other grounds of challenge to the decision of 9^th December 2010 to approve the earlier agreement

14. In order to understand the Claimant's other grounds, it is necessary to say a little more about the evidence of what took place.
15. On 7^th December 2010, Mr Hartnett wrote to Christopher Davidson who was the leader of the Large Business Service's Financial Sector part of HMRC and who had also been present at the meeting with Goldman Sachs on 19^th November 2010. The email said (I have added footnotes to explain some of Mr Hartnett's references):

"Chris can you let me know what, if anything, is wrong with the following analysis please.

When we met GS^[1] on 19/11 LBS^[2] laid out the issues under consideration. No indication was given to GS or me that the issues could not be resolved there and then from an HMRC/Business Tax/LBS perspective. GS said any settlement would have to be cleared by Esther in the States. Steve Bunson as global head of tax had flown in from NY to keep Mike^[3] under control and to settle matters if possible.

There was no suggestion that GS was within HRCP^[4] or subject to HRCP processes.

You and I at least are very familiar with the HRCP processes.

Each issue was considered individually. On some issues GS were much stronger than Richard, you or I had previously realised.

When we looked at the NICs issue I remembered that there had been a significant weakness in our technical position on interest. Mike acknowledged that and you recalled the issue but none of us could remember the detail. I recall (now) a major firm of accountants saying we had been very commercial with the settlement proposal but they had not spotted the weakness. Steve suggested that we move on interest in return for them conceding 100% of NICs. That is what we ended up doing. I am looking for the 2005 legal advice.

Richard, you and I withdrew to consider a settlement proposal and came up with proposals which I called 7 items. LBS did not mention need for governance review (which Melanie^[5] and I always do even when we are settling HRCP cases as Commissioners).

Steve said he would recommend that GS sign up to Code^[6] but this was in no sense part of a package.

Early the following week you spoke to Freda^[7] about the possible need for a governance process and later you told me you had done so and that you saw going through HRCP governance as something of a formality. GS signed up to the code on a conditional basis which at my request you asked them to make unconditional. They did so. GS were not told there was no guarantee matters could be settled.

On 26 November Freda told me that HRCP governance was being extended to cases involving £100 million TUC. That seemed very sensible. Did Freda know that the settlement with GS had been without reservation?

The HRCP programme board rejected the planned settlement of the NICs issue on the basis that interest should have been charged from 2005. You spoke to MH^[8] at GS who went off the deep end at the suggestion they should pay interest. He left me a message - I was in India – alleging extreme bad faith on your part. He repeated this when I spoke to him on Monday.

I have asked Anthony Inglese^[9] to look at any legal issues here, particularly the 2005 advice.

My concern is not so much the decision of the HRCP programme board but rather the referral after GS had been made a without reservation offer. The technical issue on interest is yet to be resolved but if we were mistaken there is precedent for HMRC accepting an HRCP settlement offer which included "relief" for a mistake by a senior member of the Dept.

I will talk to MH on Friday with a view to creating an opportunity for you to talk to him, if that is possible. If that gets us nowhere I will speak to Steve and, if necessary, see him in the US when I am there in early January.

The risks here are major embarrassment to the ChX^[10], HMRC, the LBS, you and me; not least if GS withdraw from the Code."

16. In his witness statement, Mr Hartnett says that over the next few days, he had a number of brief discussions with Ms Dawes and Ms Chaloner. Goldman Sachs had made it clear that they regarded the agreement reached on 19^th November as concluded and would oppose any attempt by HMRC to resile from it.
17. On 8^th December 2010 there was a meeting in the offices of Anthony Inglese, (Solicitor and General Counsel to HMRC). The others present were, it seems, other HMRC lawyers. Mr Hartnett was not there. There was concern among this group about a settlement with Goldman Sachs which omitted interest, in particular whether this was consistent with the Litigation and Settlement Strategy and whether it was right to impose no cost on Goldman Sachs for having resisted paying NICs so much longer than other companies who had adopted the same arrangement. Mr Inglese is recorded as saying,

"he would always want to assist [Mr Hartnett], but not if this were 'unconscionable'. He referred to the difficulty all those present at this meeting were having in justifying a settlement without an interest element."

18. Although Mr Hartnett was not at this meeting, he was advised by Mr Inglese that the Revenue could, as a matter of law, either ask Goldman Sachs to pay interest on the NICs or not do so. Mr Hartnett told the Public Accounts Committee that he was given this advice by Mr Inglese and this is also mentioned in the National Audit Office Report. It is clear from that Report that Sir Andrew Park took a different view from Mr Inglese. Sir Andrew thought the agreement which was reached on 19^th November was legally binding on HMRC.
19. In his witness statement, Mr Hartnett turned to the meeting on 9^th December 2010. He says that he considered the overall result of the settlement with Goldman Sachs was a very good one for HMRC and taxpayers generally. Goldman Sachs was maintaining that the wrong company had been assessed for NICs. They had lost in the First tier Tribunal but were appealing. It had been a good result to persuade Goldman Sachs to agree to one of the other outstanding issues. Mr Hartnett says that he was concerned that if the 19^th November 2010 settlement was re-opened there might have been a significant financial cost to HMRC.
20. HMRC's relationship with Goldman Sachs had also been uneasy. The Revenue had worked hard to improve it. He was worried that going back on the 19^th November agreement would significantly damage the relationship. That could lead to Goldman Sachs being more aggressive in their relationship with HMRC and lower tax payments in the future.
21. In December 2009 the government had published a Code of Practice on Taxation for Banks. Its objective was to encourage compliance with the spirit as well as the letter of the law. Goldman Sachs had recently adopted the Code. HMRC regarded this as a valuable step. Goldman Sachs had threatened to withdraw from the Code if HMRC withdrew from the November settlement.
22. Mr Hartnett's witness statement says,

"Additionally, and without affecting the free-standing force of our concern about the effects of Goldman's withdrawal from the code in terms of their future tax behaviour, I was concerned that withdrawal would have embarrassed the Chancellor, who had announced on 30^th November 2010 that the top 15 banks including Goldmans had signed up to the Code."

23. Further, Mr Hartnett says, he was concerned as to the impact more widely on HMRC's reputation if they withdrew from this agreement. Large business customers placed a high value on certainty and trust and it would have been unfortunate and embarrassing for HMRC to get involved in a dispute which centred on a claim that they had breached a promise.
24. By the time Mr Hartnett made his witness statement, he was aware that one of the Claimant's grounds for seeking judicial review was (and is) that he took into account an irrelevant consideration, namely embarrassment to himself and other HMRC officials. In his witness statement, Mr Hartnett responded as follows:

"For completeness, and in view of UK Uncut's grounds for judicial review, I emphasise that my own personal reputation played no part in our decision. While we discussed our disappointment that individuals within HMRC had overlooked governance processes, the embarrassment and reputation of those individuals was not something we considered relevant to our decision whether to stick with the 19 November settlement. As I have said, we were concerned about the mistake made by HMRC in relation to governance and Goldmans' justifiable perception of bad faith on the part of HMRC. Throughout my career the general approach of UK tax administration has been to honour its settlements even where a mistake has been made. This particular settlement had been agreed between senior people within HMRC and Goldmans, who could justifiably expect HMRC to comply with that settlement agreement. For the reasons I have explained we considered that it would not have been in the broader interests of HMRC and taxpayers generally to reject the settlement."

25. Ms Dawes has also provided a witness statement. She had not been involved with Goldman Sachs' tax affairs until late November 2010. She refers to her meeting with Mr Hartnett and Ms Chaloner on 9^th December 2010. She says they understood that it was open to them to let the settlement stand, or to reject it, depending on their judgment of the balance of interests under HMRC's collection and management powers. She says there were precedents with the Revenue administration of standing by agreements where the department had made a mistake (though, she adds, mistakes of this nature were rare). She says that they were firmly of the view that, had they withdrawn from the agreement, they would have had a legal challenge on their hands which they might not have won. HMRC would have lost the benefit of the settlement on the other issues and of the start of a better working relationship. They decided therefore to approve the settlement.
26. She addresses the contention in the Claimant's grounds that at this meeting she and Mr Hartnett impermissibly took account of reputational risks. She responded by saying,

"HMRC's relationship and reputation with Goldman Sachs was a key consideration for the reasons I have explained above. I would also have been conscious that going back on the settlement and provoking a dispute with Goldman Sachs had the potential to damage HMRC's reputation more widely thereby making other customers more reluctant to reach agreement with HMRC. The 'reputation and standing of officials' of HMRC was not a factor we considered when reaching our decision."

27. As I have said, the National Audit Office Report was published in June 2012. It recorded that Sir Andrew Park had been commissioned to look at 5 large settlements (of which, we now know, Goldman Sachs was one). Sir Andrew was asked to consider in each case: whether the settlement was reasonable, whether the Department's Litigation and Settlement Strategy had been followed, whether the Department had obtained appropriate legal advice and acted upon it, and whether the Department had followed its own procedures. He reviewed the Department's case papers for the settlement. He was able to submit written questions and requests for clarification. He was content that he received full answers to all of these questions and that he had had access to all the papers he considered necessary to conclude the review. He took account of the concerns expressed by whistleblowers, but had access to a greater amount of information than they had had.
28. Because of taxpayer confidentiality, the NAO report does not include the full text of Sir Andrew Park's report, but his findings are summarised. So far as the Goldman Sachs settlement was concerned, his conclusion was that,

"the overall settlement reached was reasonable considering all the circumstances. Had the only issue been settling the employer's NICs liability and charging related interest, he would not have viewed the settlement as a reasonable one. However, when the settlement is viewed as one settlement covering six issues, including the interest issue, it was reasonable. The alternative to reaching a settlement would have been to litigate all the issues. Sir Andrew Park noted that the Department might have secured a better outcome had all the issues gone to litigation, but equally, they might have done worse. He also cites the other benefits of reaching a settlement, which cannot be evaluated in monetary terms but are nevertheless important. These include normalising the relationship between [Goldman Sachs] and the Department."

29. Sir Andrew noted that of the five issues where specific sums of money were involved, the Department succeeded on three and Goldman Sachs paid in full on those matters. The two remaining issues where specific sums were involved were the waiver of interest on the NICs and a technical issue. The outcome of the technical issue was difficult to predict if it had been litigated. The sixth issue concerned management structure and Sir Andrew did not think the Department would have succeeded if that had been contested.
30. That left the question of interest. Sir Andrew considered that it was technically possible for the Department to agree not to charge interest, even when it was to receive 100% of the principal sums due. As he observed, "This is what happened when the other companies settled [i.e. in 2005]." If the issue of interest was viewed in isolation, it would have been very difficult for HMRC to justify waiving it, considering the settlement 5 years previously with the other companies who had adopted a similar arrangement. However, viewed as part of a wider settlement of the other tax issues, it was different and the overall settlement was reasonable. The report had said earlier that, in considering whether a settlement was reasonable, the primary concern was whether it had been too favourable to the taxpayer and, consequently, whether the Exchequer had received less than it should have.
31. There were the errors to which I have already referred. First, the officials at the meeting which negotiated the settlement with Goldman Sachs believed there was a barrier to charging interest. Sir Andrew Park's view was that there was no such barrier. Second, the negotiating officials did not realise that the Programme Board's approval would be necessary and therefore did not mention the need for such approval to Goldman Sachs. Sir Andrew reports that the Department took advice from the Solicitor's Office on whether the negotiated agreement was binding. The Solicitor's advice was that it was legally permissible for the Department to ask the company for interest or not to. Sir Andrew Park's view was that the agreement reached at the meeting on 19^th November was almost certainly legally binding as the settlement had been with three of the Department's top officials and the company was given no indication that further approval was necessary. His view was that, if HMRC had resiled from the agreement, there may well have been legal proceedings and the Department was right not to reopen the settlement previously negotiated.
32. So far as compatibility of the settlement with the Litigation and Settlement Strategy, the NAO report gives what it describes as its overall conclusions, based on both Sir Andrew Park's findings and those from its own work. It says that none of the settlements examined in the report (and it will be recalled that Goldman Sachs was just one of 5) involved package deals as defined in the Strategy since,

"2.17 All the disputed tax issues in each settlement were settled on their own terms, and there were no payments that were not allocated to a specific issue.

2.18 The definition of a package deal in the Litigation and Settlement Strategy does not prohibit settling an issue as part of a settlement of multiple issues on different terms than would be considered if the issue was settled by itself. For example, the settlement with [Goldman Sachs] is reasonable only in the context of being a settlement of several issues at once. Sir Andrew Park considered that had the National Insurance Contributions (NICs) and interest been the only issues, the settlement would not have been reasonable.

2.19 The Department updated the Litigation and Settlement Strategy in July 2011, and added the requirement that each disputed issue should be considered on its own merits. The guidance accompanying the 2011 version of the strategy explains that this rules out the Department conceding one issue in return for the taxpayer conceding another. This fits better with an external perspective of a package deal. However, the reality is that once the Department and a taxpayer enter a process to resolve multiple issues at once, interdependency is created between these issues and the outcome for individual issues may be different when settled in a package with other issues.

Splitting the difference

2.19 …In the settlement with [Goldman Sachs], Sir Andrew Park found that the interest on employer's NICs could be viewed as a separate tax issue so settling without the interest was not splitting the difference."

33. I can now return to summarise the Claimant's grounds for challenging the two decisions (i.e. 19^th November and 9^th December) apart from the grounds which rely on the Litigation and Settlement Strategy

i) In approving the settlement on 9^th December Mr Hartnett took account of immaterial considerations, namely:-

a) the professional embarrassment of himself and the other officials at the 19^th November 2010 because of the mistakes they had made;

b) the embarrassment which would have been caused to the Chancellor if Goldman Sachs withdrew from the Tax Code because he had recently praised the agreement of 10 big banks (including Goldman Sachs) to sign up to the Code;

c) the damage to HMRC's relationship with Goldman Sachs if the 19^th November agreement was re-opened;

This is said to be irrelevant because Goldman Sachs should not be entitled to benefit from what was said to be a previously difficult and aggressive relationship with the Revenue.

d) the damage to HMRC's reputation in the wider community if the 19^th November agreement was reopened;

This is said to be irrelevant because large firms and tax professionals would already know that the Programme Board had to approve large settlements and there would be no surprise that HMRC wished to correct errors, collect interest and comply with the Litigation and Settlement Strategy. Thus re-opening the agreement would be likely to enhance rather than damage HMRC's reputation. In any case, other large firms would know that the doctrine of legitimate expectations would provide legal backing to promises made by the Revenue if there was no good reason to resile from them.

e) Goldman Sachs' threat to withdraw from the Tax Code.

This was irrelevant since, [as Mr Hartnett recorded in his email], signing up to the Code was in no sense part of a package. Other banks could not expect favourable treatment as a result of agreeing to abide by a voluntary and unenforceable statement of general principles.

ii) Because the settlement was not a proper exercise of HMRC's functions, there was a breach of its duty under s.5(1) of the Commissioners for Revenue and Customs Act 2005 placing a responsibility on it for the collection and management of the revenue.

iii) There was a breach of the principle of equality. HMRC failed to treat like cases alike when comparison is made between this case and those of the other companies who adopted the same arrangement but who settled with the Revenue in 2005.

The Defendants' response

34. Mr Eadie's response on behalf of the Defendants was in summary as follows:

i) The 19^th November agreement did not infringe the Litigation and Settlement Strategy, as Sir Andrew Park and the NAO concluded. In any case, the Strategy provided general principles rather than bright line rules. It was not appropriate to consider it like a statute.

ii) In any case, the 19^th November agreement was overtaken by the decision on 9^th December. The situation confronting the Commissioners by that stage was out of the ordinary. It was not one to which the Litigation and Settlement Strategy had any application.

iii) Furthermore, it was necessary to have regard to the nature of this litigation. This was not a situation where a taxpayer, who had negotiated with the Revenue on the basis that it would abide by its published policy, complained of a departure. The Claimant's interest in the application for judicial review was that of taxpayers generally. They would have been affected by a departure from the policy if it meant that the Revenue collected less tax as a result and a greater burden was thereby cast on the general body of taxpayers. However, the decision on 9^th December was that overall it was more advantageous to adhere to the 19^th November agreement rather than to resile from it. Sir Andrew Park and the NAO considered that the 19^th November agreement was a reasonable one and the decision to endorse it was right. Ms Simler QC, on behalf of the Claimant, had specifically said in the course of the hearing that there was not a rationality challenge to the Commissioners' judgment. That being so, I could not conclude that the general body of taxpayers had been harmed by these agreements.

iv) Mr Hartnett and Ms Dawes specifically denied in their witness statements that the 9^th December decision had taken into account the professional embarrassment of Mr Hartnett or any other HMRC official. There had been no application to cross examine either Mr Hartnett or Ms Dawes. In those circumstances, I should proceed on the basis that their evidence was correct.

v) The Defendants accepted that any embarrassment to the Chancellor of the Exchequer was not a matter which should have been taken into account by Mr Hartnett. However, his witness statement made clear that this was an independent reason for the decision which was made on 9^th December. Ms Dawes made no mention of it. Disregarding that factor, Sir Andrew Park and the NAO considered that the 9^th December decision was the right one. On all the evidence, I should conclude that the decision would inevitably have been the same even without this immaterial consideration.

vi) As for the damage to the Revenue's relationship with Goldman Sachs if the 19^th November agreement was not ratified, the consequences of Goldman Sachs withdrawing from the Tax Code and the wider reputational damage to HMRC if it tried to withdraw from the 19^th November agreement, these were all matters for the Commissioners to weigh in their discretion. It was a general principle of administrative law that (save where Parliament prescribed certain matters which either had to be or could not be taken into consideration) it was for the decision-maker to decide what was a relevant consideration. The inclusion of a factor in the decision-making process would only be unlawful if it was irrational for it to be regarded as relevant. I could not conclude that was the case with any of these matters which the Claimant criticised the Commissioners for taking into account.

vii) Parliament had given the Commissioners a discretion as to how to best manage the collection of taxes. The Courts had interpreted this as a very broad discretion. There was no illegality in the decision to proceed with the settlement with Goldman Sachs.

viii) The principle of equality operated at a high level of generality. In particular, it could give rise to no illegality unless, at least, the two cases said to have been treated differently were materially the same and there was no other justification for different treatment. The position of Goldman Sachs in 2010 was simply different from that of the other taxpaying companies who settled in 2005.

Discussion

Litigation and Settlement Strategy grounds:

35. I will start with the Claimant's complaint that the 19^th November agreement breached the Litigation and Settlement Strategy.
36. For what it is worth, I agree with Sir Andrew Park that the agreement did not infringe the Strategy. The agreement did look at each issue separately. There was not a single payment undifferentiated as between different issues. Interest and principal were treated as separate issues. While that may seem curious, in the minds of the HMRC representatives, there was a barrier to recovering interest which did not apply to the principal. That was a mistake, but, as Sir Andrew said, it supported the idea that they were treated separately and not parts of a composite single issue. There was also a precedent for this. In 2005, the Revenue had settled with the other companies for payment of all of the outstanding NICs. Interest was waived. On this occasion, an all-or-nothing approach was taken to each of the issues: the Revenue got all of the principal of the outstanding NICs; it got nothing of the interest.
37. I have said that this is my view, for what it is worth. That is for two reasons. First, there were far more fundamental problems with the 19^th November agreement. The representatives of HMRC had not consulted with the lawyers litigating the NICs issue against Goldman Sachs and (probably as a result) they misunderstood the legal position regarding interest: there was no barrier to its recovery. In addition, they had overlooked the need to obtain approval from the Programme Board. Second, because of these errors there needed to be a further decision by the Commissioners as to whether to proceed with the settlement. Thus, it was not the decision on 19^th November which was operative, but that of 9^th December.
38. I appreciate that the Claimant submits that 9^th December decision was also tainted because neither Mr Hartnett nor Ms Dawes took account of the Litigation and Settlement Strategy. However, I agree with Mr Eadie that by this stage the Commissioners were dealing with an exceptional situation for which the Strategy was not designed. There was a dispute between the parties as to whether I should infer that Mr Hartnett and Ms Dawes must have had it in mind. Mr Eadie argued that this was a proper inference because the Litigation and Settlement Strategy was central to the way in which the Revenue conducted litigation settlements. Against that the Claimant is entitled to observe that neither Mr Hartnett nor Ms Dawes mention it in their witness statements as something they took into account on 9^th December. But it is unnecessary for me to resolve this matter. As I have said, the Strategy was beside the point. It simply did not have in contemplation a situation where an agreement had apparently been made, but where the HMRC had made some important mistakes. The issues which Mr Hartnett and Ms Dawes had to consider at least included ones which would not arise in the usual negotiation situation on which the Strategy was premised.
39. There are other reasons why I would not anyway have been inclined to grant a declaration because of the alleged errors in relation to the Strategy.
40. First, after these events the Strategy was amended. Whether the Revenue had or had not complied with the terms of the 2007 version is now of historical interest only. It is the 2011 version which will apply in the future.
41. Secondly, the Commissioners believed that the agreement of 19^th November taken as a whole represented a good deal for the Revenue. Sir Andrew Park agreed with that assessment. The Claimant does not contend that the judgment of the Commissioners in this regard is irrational, or that the 19^th November agreement did not give taxpayers value for money. As the House of Lords explained in the R v Inland Revenue Commissioners ex parte National Federation of Self-employed and Small Business Ltd [1982] AC 617 (usually known as the "Fleet Street Casuals" case), the reason why other taxpayers may have sufficient interest to challenge the decision by the Revenue to waive tax in certain situations is because an over-generous approach to one taxpayer may correspondingly increase the burden on all other taxpayers. Yet, if the 19^th November agreement was a good deal for the Revenue, that will not have been the consequence in this case.
42. At one point in her submissions, Ms Simler hinted that I should give no or little weight to the conclusions of Sir Andrew Park. She referred me to an email from Mr Hartnett to Anthony Inglese and others of 15^th December 2011 in which he reported on a conversation he had had with Amyas Morse, the Comptroller and Auditor-General, about the review which Mr Morse had commissioned Sir Andrew to carry out. In the course of this Mr Hartnett said,

"In summary, the review will focus on three issues: the reasonableness of the settlement, whether we followed our own rules, and whether lawyers were used appropriately. Amyas reminded me that he himself was at one time a tax partner (as an accountant) and dealt with many significant issues without legal support. I reminded him that in our organisation most tax work was undertaken by 'tax inspectors' and other tax officials, many of whom had a more detailed knowledge of tax provisions than some of our lawyers although that was not to be taken as any criticism of our lawyers. Amyas has been insistent that Andrew Park tackle the cases one at a time. He also told me that he has made clear to the NAO that his expectation is that nothing of substance will be found in the review." [emphasis added]

43. Ms Simler suggested that the sentence which I have emphasised was a clear indication by Mr Morse to his staff that nothing of substance should be found. She submitted that this was why the email was marked "Personal and Confidential – Please do not pass on to anyone without coming back to me".
44. I reject this argument. The sentence to which she refers is more likely to be Mr Morse's prediction as to what the review will conclude. Read in that way, it is consistent with the beginning of the paragraph in which Mr Morse referred to his experience as an accountant of dealing with significant issues without legal support. The warning about treating the email as confidential is too thin a basis for a more sinister reading of this sentence. That request is consistent with other parts of the email which could have been regarded as confidential. So, for instance, it appears that at the date it was sent, the appointment of Sir Andrew Park to this role had not been announced, the email included the names of sources on whom Mr Morse had said the Public Accounts Committee were relying (those names have been redacted in the copy of the email supplied to the Court), there was reference to discussions which Mr Morse had had with the chair of the Public Accounts Committee and to the sympathy which he had for "all the individuals in HMRC who had been bruised by the actions of the PAC and for the difficulties which the Department had experienced".

Irrelevant consideration: Mr Hartnett's personal embarrassment

45. Mr Eadie did not suggest that the decision of 9^th December could properly have been influenced by a desire to save embarrassment to Mr Hartnett or any other HMRC official. This is plainly right. Mr Eadie's case was that, on the evidence, I could not find that it did play a part. The witness statements of Mr Hartnett and Ms Dawes expressly deny that it did. Ms Simler submits that a more reliable indicator is the near contemporaneous email from Mr Hartnett of 7^th December 2010 in which he said,

"The risks here are major embarrassment to the ChX, HMRC, the LBS, you and me, not least if GS withdraw from the Code."

She argues that Mr Hartnett and Mr Davidson (to whom the email had been addressed) had made major errors in the meeting on 19^th November and Mr Hartnett was quite right that resiling from the agreement with Goldman Sachs would have been a major embarrassment for all the people he mentioned. It was inevitable that this would feature in his thinking on 9^th December when he had to decide whether or not to proceed with the agreement.

46. In my judgment, though, the Claimant could only succeed on this ground of challenge if it had given Mr Hartnett the opportunity to respond in cross examination. Oral evidence is unusual in judicial review, but it can take place if, exceptionally, the resolution of factual questions is necessary to decide the question of law and oral testimony is necessary for that purpose. It has to be Ms Simler's case that Mr Hartnett's witness statement which said "the embarrassment and reputation of those individuals was not something we considered relevant to our decision whether to stick with the 19 November settlement" was untrue. Fairness required Mr Hartnett to have the opportunity to answer that allegation orally. In circumstances such as these, in the absence of cross examination, I could not reach the conclusion which Ms Simler invites. As Stanley Burnton J. said in R (S) v Airedale NHS Trust [2002] EWHC 1780 (Admin) at [18],

"It is a convention of our litigation that in general the evidence of a witness is accepted unless he is cross examined and is thus given the opportunity to rebut the allegations made against him. There may be an exception where there is undisputed objective evidence inconsistent with that of the witness that cannot sensibly be explained away (in other words the witness's evidence is manifestly wrong)… The general rule applies as much in judicial review proceedings as in other litigation, although in judicial review, it is relatively unusual for there to be cross examination of witnesses."

The S case had a convoluted history. It was linked with the case of Munjaz and the Court of Appeal allowed the appeal ([2004] QB 395), but a further appeal by Munjaz was also successful ([2006] 2 AC 148), but so far as I can see, neither of the higher courts criticised the passage which I have quoted above. In any case, views to a similar effect were expressed by the Court of Appeal in R (MWH and H Ward Estates Ltd) v Monmouthshire County Council [2002] EWCA Civ 1915 at [29].

47. When this judgment was distributed in draft, the Claimant commented that S v Airedale NHS Trust had not been referred to in the hearing. That was correct. The Claimant made brief written submissions in relation to that case and the Defendant made written submissions in reply. I have taken both into account. The Claimant did not seek to argue that the observations of Stanley Burnton J. were wrong. Rather, the Claimant noted that, on the facts of the case before him, he had found that there was a contemporaneous document which was inconsistent with the witness statement and he preferred the evidence of the contemporaneous document even in the absence of cross examination. The Claimant argued that, likewise, the email of 7^th December was a contemporaneous document which was inconsistent with Mr Hartnell's witness statement and the email should be preferred. The Claimant argued that it was significant that the witness statement did not address the email and submitted no further witness statement when that point was taken by the Claimant in its skeleton argument for the hearing. In the alternative, the Claimant submitted, I should resume the hearing and take oral evidence from Mr Hartnett.
48. The Defendant responded by saying that the facts of the S case were different to the present one. The 7^th December email did not purport to set out the reasons for the 9^th December decision, nor was it contemporaneous with it. These circumstances did not come within the exception which Stanley Burnton J. articulated in S or found applicable to the facts of that case. Fairness to the Claimant did not require the hearing to be resumed. The Defendant had relied on the authority of MHW and H Ward Estates for a similar proposition to that in S. The Claimant elected not to apply to cross examine Mr Hartnett, but instead invited the Court to reject his witness statement. Fairness did not require the Claimant to be given a further opportunity. On the contrary it would involve a waste of public funds.
49. I agree with the Defendant, for the reasons that they give, that the Claimant is not helped by the different outcome of the S case. The passage which I quoted from the decision of Stanley Burnton J. acknowledged that there may be circumstances where a court can, and should, accept that a witness statement is manifestly wrong because it is contradicted by contemporaneous evidence. The email of 7^th December 2010 may have given the Claimant ammunition to argue that, unusually, Mr Hartnett should be cross examined on his witness statement. But the test which Stanley Burnton J. set for rejecting evidence in a witness statement without cross examination is more stringent than that which would have been applied on such an application.
50. I also reject the submission that the hearing should be resumed to allow such cross examination to take place. The Defendant cited MHW and H Ward Estates in their skeleton argument (where the significance of the absence of cross examination was made). In the course of the hearing, I also confirmed with Ms Simler that there had been no application for cross examination. The Claimant chose to proceed on the material available to it. Fairness does not require the Claimant to be able to have a second go.
51. For these reasons and despite the Claimant's further submissions, my view remains the same. The "personal embarrassment" ground does not succeed.

Irrelevant consideration: embarrassment to the Chancellor

52. The Defendants accept that this was an irrelevant consideration for Mr Hartnett to take into account. However, they submit that it was not a factor which influenced Ms Dawes and, in any case, it is plain that the decision would have been the same even if Mr Hartnett had ignored the impact on the Chancellor of Goldman Sachs withdrawing from the Tax Code.
53. I am not persuaded by Mr Eadie's first response. If one of two decision makers takes account of an irrelevant matter that is, in principle at least, sufficient to render the decision unlawful. The Commissioners empowered any two of their number to take decisions on their behalf. In those circumstances I consider that both of them had to take account of only relevant considerations.
54. As to the second, Ms Simler reminded me of the difficulty of disentangling one factor from others, and in R (FDA) v Secretary of State for Work and Pensions [2013] 1 WLR 444, Lord Neuberger summarised the principles at [67] – [68]. He said,

"67. Where a decision-maker has taken a legally irrelevant factor into account when making his decision, the normal principle is that the decision is liable to be invalid unless the factor played no significant part in the decision-making exercise…

68. Even where the irrelevant factor played a significant or substantial part in the decision-maker's thinking, the decision may, exceptionally, still be upheld, provided the court is satisfied that it is clear that, even without the irrelevant factor, the decision-maker would have reached the same conclusion. Thus in the Simplex case 57 P&CR 306, 326, Purchas LJ approved the following passage in the judgment of May LJ in R v Broadcasting Complaints Commission ex parte Owen [1985] QB 1153, 1177,

'Where the reasons given by a statutory body for taking … a particular course of action are not mixed and can clearly be disentangled, but where the court is quite satisfied that even though one reason may be bad in law, nevertheless the statutory body would have reached precisely the same decision on the other valid reasons, then this court will not interfere by way of judicial review.'

In R (Smith) v North Eastern Derbyshire Primary Care Trust [2006] 1 WLR 3315, [10] (a different) May LJ said this:

'Probability is not enough. The defendant would have to show that the decision would inevitably have been the same and the court must not unconsciously stray from its proper province of reviewing the propriety of the decision-making process into the forbidden territory of evaluating the substantial merits of the decision (emphasis added).'"

55. In this case, Mr Hartnett says in his witness statement that his wish to avoid embarrassment to the Chancellor was an additional and distinct reason for his decision to approve the 19^th November agreement.
56. Ms Simler took me to the judgment of Stanley Burnton J. in R (Nash) v Chelsea College of Art and Design [2001] EWHC 538 (Admin) at [34] for the principles concerning late reasons by decision-makers. Mr Eadie is right to comment that these principles are not entirely on point. This is not a case where Mr Hartnett and Ms Dawes were under any legal duty at the time to give reasons for their decision on 9^th December 2010. However, the warning is still apposite in the sense that a Court needs to be cautious of later reasons and be aware of the risk that they have been composed subsequently to justify the decision and are a retrospective justification of that original decision.
57. In this case, though, I do conclude that the decision would have been the same even if Mr Hartnett had not taken account of potential embarrassment to the Chancellor. He gives other independent and substantial reasons why the 9^th December decision was taken. That those would, separately from the irrelevant consideration, have led to the same decision is supported by the fact that Ms Dawes reached her decision without regard to it. That those other reasons were significant and substantial reasons for taking the same decision is supported by the report of Sir Andrew Park and the NAO.
58. Furthermore, Ms Simler said that the purpose of seeking a declaration, even without a quashing order, was so that HMRC would learn the errors of its ways. But a declaration is unnecessary to achieve that aim in the present context. HMRC accepts that potential embarrassment to the Chancellor was an irrelevant consideration. Mr Hartnett should not have taken it into account. In those circumstances a declaration would serve no purpose.

Irrelevant considerations: damage to the relationship with Goldman Sachs, damage if Goldman Sachs withdrew from the Tax Code, damage to the wider reputation of HMRC

59. The Commissioners for Revenue and Customs Act 2005 s.5(1) says that the Commissioners shall be responsible for

"(a) the collection and management of revenue for which the Commissioners of Inland Revenue were responsible before the commencement of this section."

"Revenue" includes National Insurance Contributions – see s.5(4).

60. In the Fleet Street Casuals case the House of Lords had to consider an earlier generation of Revenue legislation, but the Claimant did not suggest it was materially different. As Lord Diplock said at p. 636,

"In the exercise of these functions the board have a wide managerial discretion as to the best means of obtaining for the national exchequer from the taxes committed to their charge, the highest net return that is practicable having regard to the staff available to them and the cost of collection."

61. The Fleet Street Casuals case was decided at a relatively early stage in the modern development of judicial review, but the principle enunciated by Lord Diplock has not changed. In R (Davies) v HMRC; R (Gaines-Cooper) v HMRC [2011] 1 WLR 2625 at [26] Lord Wilson said,

"The primary duty of the revenue is to collect taxes which are properly payable in accordance with the current legislation but it is also responsible for managing the tax system: section 1 of the Taxes Management Act 1970. Inherent in the duty of management is a wide discretion. Although the discretion is bounded by the primary duty (R (Wilkinson) v Inland Revenue Commissioners [2005] 1 WLR 1718 [21] per Lord Hoffman), it is lawful for the Revenue to make concessions in relation to individual cases or types of case which will, or may, result in the non-collection of tax lawfully due provided that they are made with a view to obtaining overall for the national exchequer the highest net practicable return: R v Inland Revenue Commissioners v National Federation of Self-Employed and Small Businesses Ltd [1982] AC 617, 636 per Lord Diplock. In particular, the revenue is entitled to apply a cost-benefit analysis to its duty of management and in particular, against the return thereby likely to be foregone, to weigh the costs which it would be likely to save as a result of a concession which cuts away an area of complexity or likely dispute."

62. Mr Eadie's proposition that it is for the decision-maker to decide what factors are relevant or irrelevant, subject, of course, to any statutory requirement and Wednesbury irrationality, is supported by R (Khatun) v Newham LBC [2005] QB 37 especially what Laws LJ said at [35]. He was there considering an argument that a decision-maker was obliged to treat certain factors as relevant. But exactly the same reasoning must apply to a claim that a decision-maker was obliged to disregard particular factors as irrelevant. Absent a statutory injunction to disregard the factor in question, it is for the decision-maker himself or herself to decide what to take into account, subject only to a rationality challenge.
63. In my judgment it was not irrational for Mr Hartnett and Ms. Dawes to take into account the factors on which the Claimant relies and which I am discussing under this heading. Whether or not the advantage of a better relationship between Goldman Sachs and the Commissioners, whether or not the advantage of Goldman Sachs continuing to sign up to the Tax Code, whether or not the danger of undermining HMRC's reputation for abiding by agreements even if they were based on mistakes were each of sufficient significance that they needed to be taken into account in deciding whether to endorse the 19^th November agreement were quintessentially questions to be decided by the Commissioners themselves within the broad managerial discretion given to them by statute. In my judgment the Claimant cannot show that it was irrational for the Commissioners to take any of them into account. Accordingly, this part of the challenge fails.

Breach of duty under Commissioners for Revenue and Customs Act 2005 s.5

64. This challenge is parasitic on the others. The Claimant argues that because of the other legal flaws in their decision making process the Commissioners have not complied with their duty to collect and manage taxes and NICs. Since those other challenges have not been made out, this challenge also fails.

Principle of equality

65. It is fair to say that this did not feature prominently in Ms Simler's submissions. It is sufficient for me to say that I agree with Mr Eadie's response. There was not such a close parallel between Goldman Sachs and the companies who settled in 2005 that HMRC were obliged to treat them in the same way (or, more strictly, charge Goldman Sachs interest on the unpaid NICs between 2005 and 2010). As Sir Andrew Park said, there would have been force to that point if the issue of unpaid NICs and interest was considered in isolation, but HMRC was entitled to consider that the settlement of all the outstanding issues with Goldman Sachs justified the agreement as a whole. That was the position as at 19^th November. For the reasons I have already given, on 9^th December HMRC was entitled to consider that there were also other issues in play. As the Court of Appeal said in R (E) v Nottinghamshire Healthcare NHS Trust [2010] EWCA Civ 795 at [90] the principle of equality simply means that distinctions between different groups must be drawn on a rational basis and is thus no more than an example of the Wednesbury principle. As I have already noted, the Claimant does not mount a challenge to the HMRC decisions on the basis that they were Wednesbury irrational.

Conclusion

66. The settlement with Goldman Sachs was not a glorious episode in the history of the Revenue. The HMRC officials who negotiated it had not been briefed by the lawyers who were litigating against Goldman Sachs. They relied on their belief or recollection that there was a barrier to the recovery of interest on the unpaid NICs. That was erroneous. HMRC accepts now that there was no such barrier. The officials who negotiated the agreement overlooked the need for approval from the Programme Board in relation to an agreement over £100 million. HMRC now accepts that they should have appreciated this. Because the officials did not have this requirement to mind, they said nothing about it to Goldman Sachs and created the impression that the agreement was a done deal by the end of the meeting on 19^th November. HMRC accepts that was an error. Furthermore, HMRC did not appear to have taken a contemporaneous note as to the agreement which was reached on 19^th November. That allowed a degree of uncertainty to prevail for a time as to what precisely had been agreed. In the end that has been resolved but in the course of the hearing, HMRC accepted that it would have been a good idea for a contemporaneous record to have been kept. Next, by his own admission, when he decided to approve the settlement on 9^th December 2010, Mr Hartnett took into account the potential embarrassment to the Chancellor of the Exchequer if Goldman Sachs were to withdraw from the Tax Code. HMRC accepts that was an irrelevant consideration and should not have featured in his decision-making process.
67. However, my task is to decide whether the decisions of HMRC under challenge were unlawful. As Simon J. said when granting permission to apply for judicial review, "maladministration and illegality are separate issues." He was echoing the sentiments expressed by Lord Scarman in the Fleet Street Casuals case when he said at p. 652,

"The duty of fairness as between one taxpayer and another is clearly recognised in these (and other passages) in the modern case law. Is it a mere moral duty, a matter for policy but not a rule of law? If it be so, I do not understand why distinguished judges allow themselves to discuss the topic: they are concerned with law, not policy. And is it acceptable for the courts to leave matters of right and wrong, which give rise to genuine grievance and are justiciable in the sense that they may be decided and an effective remedy provided by the courts, to the mercy of policy? Are we in the twilight world of 'maladministration' where only Parliament and the Ombudsman may enter, or upon the commanding heights of the law? The courts have a role, long established, in public law. They are available to the citizen who has a genuine grievance if he can show that it is one in respect of which prerogative relief is appropriate. I would not be a party to the retreat of the courts from this field of public law merely because the duties upon the revenue are complex and call for management decisions in which discretion must play a significant role."

68. But it is worth remembering that Lord Scarman's rhetoric was directed at the argument that the National Federation of Self-Employed and Small Businesses Ltd's claim could not be examined because it lacked standing to apply for judicial review. When the claim was examined, he, like the other members of the House of Lords, concluded that it was not sustainable. Similarly, Mr Eadie did not suggest that I should retreat from the "commanding heights of the law", but rather I should conclude that the Claimant's arguments that HMRC had erred in law, as well as making the errors which he admitted, were not made out. For all of the reasons which I have given above, I agree.
69. Consequently, this application for judicial review is dismissed.

Note 1   Goldman Sachs    [Back]

Note 2   Large Business Service     [Back]

Note 3   Mike Housden, a representative of Goldman Sachs    [Back]

Note 4   High Risk Corporates Programme    [Back]

Note 5   Melanie Dawes, another Commissioner    [Back]

Note 6   A Code of Practice on Taxation for Banks, issued by HMRC in December 2009    [Back]

Note 7   Freda Chaloner, chair of the Programme Board    [Back]

Note 8   Mike Housden    [Back]

Note 9   Solicitor and General Counsel to HMRC    [Back]

Note 10   The Chancellor of the Exchequer    [Back]