Royal Courts of Justice
Strand, London, WC2A 2LL
Before :
THE HON. MR JUSTICE HOLGATE
Between :
THE QUEEN on the application of (1) BT Pension Scheme Trustees Limited (2) Marks and Spencer Pension Trust Limited (3) Ford Pension Scheme for Senior Staff Trustees Limited (4) Ford Salaried Pension Fund Trustees Limited (5) Ford Pension Fund Trustees Limited | Claimants |
- and – | |
(1) UK Statistics Authority (2) Chancellor of the Exchequer | Defendants |
Javan Herberg QC, Fraser Campbell and Daniel Cashman (instructed by Slaughter and May) for the Claimants
Sir James Eadie QC, Tom Hickman QC, David Lowe and Tom Leary (instructed by Government Legal Department) for the Defendants
Hearing dates: 21 and 22 June 2022
Approved Judgment
Mr. Justice Holgate:
Introduction
The central issue raised by this claim for judicial review is whether the UK Statistics Authority (“UKSA”) has acted unlawfully by deciding that the United Kingdom General Index of Retail Prices (“the RPI”) will in future import the methodology and data sources of the Consumer Prices Index including owner occupiers’ housing costs (“the CPIH”). The parties refer to this as “the RPI decision”. As explained below, the Chancellor of Exchequer has withheld his consent to that decision being implemented before 2030.
It is common ground that for reasons to do with statistical methodology, the RPI produces an estimate of inflation about one percentage point higher than the CPIH and that this will continue in the long term. The UKSA decided some years ago that the methodology of the RPI is flawed and so that index should cease to be used as the UK’s measure of inflation. Those decisions have never been the subject of legal challenge.
The five claimants are large pension funds with substantial liabilities for pension payments linked to the RPI, and in some cases the Consumer Prices Index (“the CPI”). Following good practice, they have invested heavily in investments aimed at matching those liabilities, that is index-linked Government gilts. All such gilts are index-linked to the RPI. Their market value is driven by the net present value of the future income stream they will generate. The impact of the long-run reduction of 1% in the RPI from 2030 onwards, affecting future interest payments to holders of the gilts, is said to be around £90 billion - £100 billion.
There are 10.5 million people with private sector “defined benefit” pensions, of whom a majority have pension schemes linked to the RPI (for the revaluation of “preserved benefits” or indexation of pensions in payment or both). It has been estimated that they will receive reduced payments amounting to around 4 - 9% of their lifetime benefits, with women experiencing on average a greater reduction.
The claimants say firstly, that the RPI decision involves a breach of the UKSA’s duty under s.21(1) of the Statistics and Registration Service Act 2007 (“SRSA 2007”) to compile, maintain and publish every month the RPI. Secondly, they say that the UKSA has misinterpreted the legislation by treating those impacts on pensions and pension funds as being legally irrelevant to its statutory functions.
The RPI has a long history. In 1947 the Government’s Cost of Living Advisory Committee recommended the creation of a new index of changes in household consumer prices which was initially known as the Interim Index of Retail Prices. In January 1956 the index was adopted as the official measure of UK inflation and its name was changed to the RPI.
The RPI has been used for a wide variety of purposes in addition to providing index-linking for gilts and pensions. These include price regulation (e.g. train fares and utility charges), interest rates on student loans, wage negotiations, and the adjustment of certain tax thresholds and liabilities.
The UK Government introduced index-linked gilts in 1981. They were linked to the RPI in preference to other indexes at that time, such as the earnings index, which would have significantly increased borrowing costs for the Government. At that stage the RPI was seen to be a good cost of living index, it was already used for index-linked state pensions, and it was published monthly.
Index-linked gilts issued from 1981 and before July 2002 (referred to as “old-style gilts”) contained a clause entitling a holder to redeem the gilt before it matures if any change should be made to the coverage or the basic calculation of the RPI which, in the opinion of the Bank of England, constitutes a “fundamental change” in the Index “which would be materially detrimental to the interests of stockholders”. The last of the gilts with this early redemption clause matures in 2030. It was decided that gilts issued from July 2002 onwards would not contain any such provision for early redemption, in line with the approach taken in other major markets.
Gilts issued from September 2005 onwards include a clause which, in summary, defines the Index as being the RPI, or if the “existing Index ceases to be published”, any replacement Index selected by the Chancellor as continuing “the function of being an officially recognised index measuring changes in the level of UK retail prices”. Although the main purpose of this clause is to define an index which can continue to be used to adjust the value of the gilt for calculating interest and the redemption payment, I will adopt the term which the parties used to describe it, that is “the cessation clause”.
When the SRSA 2007 established the UKSA and made it solely responsible for maintaining the RPI, s.21(3) gave the Chancellor a power to veto any decision by the Authority to make a “fundamental change” in that index qualifying as a trigger for the early redemption of an old-style gilt. Vanessa MacDougall was Director Economics and Deputy Chief Economic Adviser at HM Treasury between April 2017 and October 2020. In her witness statement she explains that this provision enabled the Chancellor to protect the public finances and the orderly functioning of the gilts market by vetoing a change which could result in early redemptions of gilts on a substantial scale (paras. 18 and 44).
In practice the Chancellor will not be able to use this power from 2030 when there will no longer be any gilts with early redemption clauses. Up until then the Chancellor’s function under the SRSA 2007 is limited to deciding whether or not to exercise this veto. The UKSA’s obligation under s.21(1) to maintain the RPI carries on beyond 2030, regardless of whether the Chancellor is able to exercise his power of veto under s.21(3).
The CPI was first published in 1997 (under the then title the “Harmonised Index of Consumer Prices”) in order to comply with EU Council Regulation (EC) No.2494/95 for the purposes of producing a consistent measure of inflation across Europe. In 2013 the Office for National Statistics (“ONS”), the executive office of the UKSA, introduced the CPIH, that is the CPI with the addition of owner occupiers’ housing costs and council tax.
Just over a decade ago there began to be serious concerns about the methodological differences between the RPI and the CPI (and subsequently the CPIH), which resulted in the former generally over-estimating inflation. In 2012 the then National Statistician (the chief executive of the UKSA) consulted on options for reforming the RPI. In January 2013 she made a public announcement that a particular formula used in the RPI did not meet international standards. She stated that the ONS would not have used that formula if it had been constructing a new index, but there was significant value in maintaining the long time series of the RPI without major change, so that it could continue to be used for long-term indexation and for index-linked gilts and bonds. Because of the flaws in the RPI, the UKSA decided in March 2013 to de-designate it as a “National Statistic” under s.14 of the SRSA 2007.
In May 2013 the UKSA asked Mr Paul Johnson, then Director of the Institute for Fiscal Studies, to carry out a review of price indexes. In January 2015 he concluded that the RPI was a flawed statistical measure of inflation which should not be used for new purposes and which should cease to be used for all purposes, except as a “legacy measure” where contractually required. He recommended that the Government and regulators should work towards ending the use of RPI as soon as practicable.
As a result, the then National Statistician wrote to the Chair of the UKSA on 9 March 2016, stating that he strongly discouraged the use of the RPI as a measure of inflation as there were far superior alternatives. The RPI would continue to exist for legacy purposes and be maintained by making only “routine changes” (e.g. updating of data), rather than “methodological changes”. He stated that the CPIH should become the ONS’s preferred measure of consumer inflation. On 31 July 2017 the CPIH was designated under the SRSA 2007 as a National Statistic.
On 12 June 2018 the House of Lords Economic Affairs Committee launched an inquiry into the use of the RPI as a measure of inflation. It published its report on 17 January 2019. The Committee criticised the UKSA for having failed to correct flaws in the RPI in accordance with its duty to promote and safeguard the quality of statistics. This had created winners and losers. The holders of RPI-linked gilts had been winners because their interest payments had increased by about £1 billion a year. But losers included commuters and students because annual rail fare increases and interest on student loans were linked to the RPI. Furthermore, the divergence between CPI and RPI had encouraged Government to “index shop”, citing the example of its decision in 2011 to switch the uprating of benefits, tax thresholds and public sector and state pensions from the RPI to the CPI. The Committee noted that despite both the recommendations of the Johnson Review and discouragement from the UKSA, the RPI remained in widespread use not only for index-linked gilts, but also corporate bonds and for uprating private sector pensions.
The Committee considered that it was untenable for the UKSA to continue to publish the RPI with its admitted methodological flaws. Pursuant to its statutory duty the Authority should promote a solution to the “formula problem” and under the SRSA 2007 the Chancellor should consent to that change in RPI methodology and not veto it, regardless of the effects on the holders of index-linked gilts, who had received an “unwarranted windfall”. “Given RPI remains in widespread use, the Authority should stop treating RPI as a legacy measure and resume a programme of periodic methodological improvements.” No doubt “widespread use” referred to the overall range of persons and bodies affected by the continued use of the RPI, and not simply the holders of index-linked financial instruments.
The Committee also recommended that the UKSA and the Government should agree on a single measure of inflation for official use within the next 5 years. In the interim the Government should switch to the CPI from the RPI in all areas where it was still being used, apart from private contracts. Upon a new single measure being agreed, the Government should begin to issue gilts linked to that index.
On 14 February 2019 the UKSA’s Board accepted the recommendation of the National Statistician that firstly, the obligation in the SRSA 2007 to maintain and publish the RPI should be repealed and secondly, in the interim, the shortcomings of the RPI should be corrected by bringing the methods and data sources of the CPIH into the RPI. The UKSA submits that that amounted to a “decision” by the Board for the purposes of making a claim for judicial review. The Claimants submit that no decision was made at that stage. The Board simply agreed to put forward a proposal. It did not make any decision in principle that the CPIH should be brought into the RPI. Either way, the parties agree that the UKSA did not consult on the principle of making this change, whether before or after February 2019.
On 18 February 2019 the Chair of the UKSA wrote to the Bank of England under s.21(2) of the SRSA 2007. In its reply of 4 March 2019 the Bank of England decided that bringing the methods of the CPIH into the RPI would involve a fundamental change to the latter which would be detrimental to the interests of the holders of gilts with redemption clauses. Consequently, on the same day the UKSA wrote to the Chancellor seeking his consent to that change under s.21(3) of the SRSA 2007.
On 4 September 2019 the Chancellor gave his formal response. He said that, because of potential disruption for users of the RPI, he would not promote legislation to repeal the requirement in s.21(1) for the RPI to be compiled and published. On the s.21(3) issue, the Chancellor acknowledged that the requirement for his consent would expire in 2030, at which point the then Board of the Authority would be likely to align the RPI with the CPIH. As for the period before 2030, the Chancellor had regard to the integrity of the statistical system, the effect on public finances and on the holders of gilts with redemption clauses. He considered that there would be significant effects for users of the RPI. However, the full extent of those effects and how users might respond or adjust contractual arrangements was not known. Given the potential for substantial and diverse effects, it was appropriate to assume that users would need a substantial period of time to prepare for any such change. Accordingly, the Chancellor said that, based on the information available to him, he could not consent to the change under s.21 any earlier than February 2025. He also said that to ensure better information about potential effects, the Government would consult publicly on whether the change should be made before 2030 and, if so, when between 2025 and 2030. As part of the consultation, the UKSA would consult on “technical matters concerning how to implement the proposed alignment of RPI with CPIH.”
On the same day, public announcements were made by the Chancellor and by the UKSA on that basis. The Authority made it plain that it would consult solely “on the method of making the change”.
HM Treasury and the UKSA issued a joint consultation document on 11 March 2020 (“A Consultation on the Reform to the Retail Prices Index Methodology”). As the claimants’ skeleton states, the consultation focused on firstly, the technical approach to be taken by the UKSA to import the methods and data sources of the CPIH into the RPI and secondly, on whether the change should be made before 2030 and, if so, when between 2025 and 2030. The lengthy consultation period ended on 21 August 2020.
On 23 October 2020 the Chancellor wrote to the Chair of the UKSA to say that, having considered the responses submitted in the consultation, he would not consent to the change to the RPI before the maturity of the final “old-style gilts” in 2030 “in order to minimise the impact of UKSA’s reform to the RPI on the holders of index-linked gilts”.
In his reply to the Chancellor dated 20 November 2020, the Chair of the UKSA said it was the Authority’s policy to address the shortcomings of the RPI in full at the earliest time, by bringing methods and data sources from the CPIH into the RPI. That change could legally and practically be made in February 2030.
HM Treasury and the UKSA published their joint response to the consultation exercise on 25 November 2020. That document announced decisions which the claimants seek to challenge as follows: -
The UKSA confirmed that from 2030 it would use the methodology proposed in the consultation document for bringing the methods and data sources of the CPIH into the RPI (“the RPI decision”);
The Chancellor withheld his consent under s.21(3) of the SRSA 2007 to this alteration of the RPI by the UKSA at any time before the last gilt with an early redemption clause matures in 2030 (“the timing decision”);
In response to representations made during the consultation, the Chancellor decided that the Government would not pay compensation to the holders of index-linked gilts (“the compensation decision”).
In summary, the claimants advance the following grounds of challenge: -
Ground 1
The decision of the UKSA that the RPI should be amended by importing the methods and data sources of the CPIH (“the RPI decision”) falls outside the scope of the power to amend under s.21(1) of the SRSA 2007 and is therefore ultra vires. Each defendant proceeded on the erroneous legal assumption that that amendment would be intra vires the SRSA 2007;
Ground 2
(1) The UKSA failed to have regard to an obviously material consideration, the impact of the decision to change the method of calculation of the RPI (the RPI decision) on legacy users, that is holders of RPI index-linked gilts and bonds and persons entitled to index-linked pensions, or wrongly decided that it was not entitled to have regard to that consideration. As a result the UKSA failed to comply with the Public Sector Equality Duty (“PSED”) under s.149 of the Equality Act 2010;
(2) In relation to the compensation decision, the Chancellor failed to have regard to the interests of legacy users and to comply with the PSED.
Ground 3
(1) The UKSA was under a duty to consult the public on its RPI decision and failed to do so. The UKSA failed to consult on the intended change to the RPI when that proposal was at a formative stage and failed conscientiously to take into account representations made by legacy users such as the claimants;
(2) The Chancellor was under an obligation to consult with legacy users on the issue of compensation and failed to do so. The Chancellor failed to give consultees any information on compensation to enable them to give due consideration and to respond on that issue. The Chancellor failed to take conscientiously into account such concerns as were raised by consultees.
It may seem odd for the claimants to be asking the court to quash the Chancellor’s timing decision, because that is favourable to them and all those with similar interests. The Chancellor refused his consent under s.21(3) to the implementation of the RPI decision at any time before 2030. Thereafter, the Chancellor has no further power to intervene under the SRSA 2007. However, the explanation for their challenge is that the timing decision assumed that the Chancellor was dealing with a change to the RPI which was intra vires s.21(1) and (3). If the claimants succeed on ground 1, they simply say that the timing decision falls away.
The claim form states that the claimants challenge decisions taken by both defendants on 25 November 2020; that is the date when the joint response to the consultation that year was published. On that basis the claim issued on 19 February 2021 was brought within the 3 months’ time limit set by CPR 54.5(1). However, the UKSA submits that the RPI decision was taken on 14 February 2019 when the Authority’s board met and that the challenge, in particular, to the scope of its consultation (ground 3(1)) is long out of time. Similarly, ground 3(1) alleges that when the consultation began in March 2020 the RPI decision had previously been taken; it was “not at a formative stage”. Plainly, the claimants cannot have it both ways. In any event, as I explain below, it is clear from the contemporaneous documents which were made available to the public on 4 September 2019, that in substance the RPI decision had been taken on 14 February 2019.
The Statement of Facts and Grounds also includes a private law claim for a declaration that, if the court should conclude that the RPI decision stands, the cessation clause in gilts issued from 2005 onwards will be triggered in 2030 by bringing into the RPI the data sources and methodology of the CPIH. The claimants argue that the consequence of that change will be that the RPI will cease to be published. The Chancellor submits that, to the contrary, the RPI will continue to be published, notwithstanding those changes. Both he and the claimants are agreed that the court should make a declaration as to which of those competing views is correct, on the material now before the court.
I am grateful for the considerable assistance I received from counsel and the parties’ legal teams as regards the written and oral submissions and the way in which the case was prepared for the hearing. The court also had the benefit of witness statements from Mr. Nicholas Groom on behalf of the first claimant, Mr. Graham Oakley on behalf of the second claimant, Mr. Sherman Garner on behalf of the third, fourth, fifth claimants, Mr. Jonathan Athow on behalf of the first defendant, and Ms. MacDougall on behalf of the second defendant. I found their evidence very helpful in understanding the factual background and the issues raised by the claim. Although I have had full regard to that material, it has not been necessary to summarise all of it in this judgment.
The remainder of this judgment is set out under the following headings: -
Heading | Paragraph numbers |
Statutory framework | 34 - 68 |
A comparison of the RPI and the CPIH | 69 - 78 |
Addressing the flaws in the RPI | 79 - 118 |
Effects on legacy users of the change to the RPI | 119 - 127 |
Ground 1 | 128 - 155 |
Ground 2 | 156 - 189 |
Ground 3 | 190 - 208 |
Delay | 209 - 215 |
The private law claim | 216 - 237 |
Statutory Framework
The regime before the Statistics and Registration Service Act 2007
The regime before the enactment of the SRSA 2007 was essentially non-statutory. The then Office for National Statistics was the central producer of statistics in the UK. It was an Executive Agency accountable to the Chancellor of the Exchequer. The relationship between the two was set out in the Framework for National Statistics (June 2000).
“National Statistics” “provide an accurate, up-to-date, comprehensive and meaningful description of the UK economy and society” “used by Government, Parliament and the wider community in decision-making and debate ….”. The concept included firstly, all the public access databases and publications produced by the ONS and secondly, many of the key public interest statistics produced by government departments and devolved administrations (paras. 1.1 to 1.2).
The National Statistician was the professional Head of National Statistics and Director of the ONS. He was responsible for ensuring that statistics were produced in accordance with the standards set out in the National Statistics Code of Practice. Like the ONS, the National Statistician was accountable to the Chancellor.
Paragraph 3.4 of the Framework stated that the National Statistician was responsible for “developing and maintaining statistical standards, definitions and classifications and promoting high quality statistical output through systematic evaluation and research”, except in relation to the RPI where “special arrangements” applied: -
“The National Statistician will take the lead in advising on methodological questions concerning the RPI but the scope and definition of the index will continue to be matters for the Chancellor of the Exchequer.”
This enabled the Chancellor, for example, to prevent any changes to the RPI he considered would be likely to have an adverse effect on public finances. It was also well-understood that changes to the RPI could have an effect in many areas of society (MacDougall WS paras. 28-29).
However, this arrangement did not accord with best international practice. No other country reserved control over a central price index to a Minister. On 18 January 2001 the Treasury Select Committee drew the clear conclusion that it was “absolutely essential” that the arrangements for National Statistics should be enshrined in legislation which would inter alia “guard against political interference”. The Royal Statistical Society and the Statistics Commission criticised the arrangements for the RPI as contributing to the perception of ongoing Ministerial interference in the production of official statistics.
In November 2005 the Chancellor announced that the Government would legislate for an independent Statistics Board with responsibility for the governance and publication of official statistics in the UK. Consultation on the proposals followed in 2006. The Bill that became the SRSA 2007 gave Parliament the opportunity to consider how the RPI should be treated under the new regime.
The Statistics and Registration Service Act 2007
Section 55 provided that both the ONS as previously constituted and the Statistics Commission should cease to exist.
Section 1 established the Statistics Board, referred to as the UKSA, to exercise functions on behalf of the Crown (s.2). The Board comprises the chairman appointed by HM The Queen and at least 5 non-executive members appointed by the Minister for the Cabinet Office. The executive members are the National Statistician and two other employees of the Board (s.3).
The UKSA is a “Non-Ministerial Department”, acting at arm’s length from Ministers. It is directly accountable to Parliament, not to any Minister (paras. 2.1 – 2.2 of the Memorandum of Understanding between the Cabinet Office and the UKSA issued in April 2020). It must produce an annual report to Parliament and the devolved assemblies each year (see s.27).
Paragraph 2.3 of the Memorandum states: -
“Ministers have no role in the decision-making of the Authority Board. Ministers’ relationship to the Board will be consistent with the Code of Practice for Statistics (including as referenced in the Ministerial Code), the UN’s Fundamental Principles of Official Statistics ….”
In relation to the National Statistician s.30 provides: -
“(1) The National Statistician is to be the Board's principal adviser on—
(a) the quality of official statistics,
(b) good practice in relation to official statistics, and
(c) the comprehensiveness of official statistics.
(2) The Board must have regard to the advice of the National Statistician on those matters.
(3) If the Board rejects the advice of the National Statistician—
(a) in relation to the development and maintenance of definitions, methodologies, classifications and standards for official statistics, or
(b) as to the application to any statistics produced by the Board of any definition, methodology, classification or standard promoted by the Board under section 9,
the Board must publish a statement of its reasons for doing so.
(4) Where the Board publishes a statement under subsection (3), it must as soon as possible thereafter lay a copy before Parliament.”
The National Statistician is appointed by the Crown (s.5(2)) and is the chief executive of the UKSA (s.31(2)). He or she must establish an executive office of the Board (s.32). That has been named the ONS.
Section 6 defines “official statistics” to include statistics produced by the UKSA, a government department or a devolved administration.
Section 7 sets out the “Objective” of the UKSA: -
“(1) In the exercise of its functions under sections 8 to 21 the Board is to have the objective of promoting and safeguarding the production and publication of official statistics that serve the public good.
(2) In subsection (1) the reference to serving the public good includes in particular—
(a) informing the public about social and economic matters, and
(b) assisting in the development and evaluation of public policy.
(3) The Board is accordingly, in the exercise of its functions under sections 8 to 21, to promote and safeguard—
(a) the quality of official statistics,
(b) good practice in relation to official statistics, and
(c) the comprehensiveness of official statistics.
(4) In this Part references to the quality of any official statistics includes—
(a) their impartiality, accuracy and relevance, and
(b) their coherence with other official statistics.
(5) In this Part references to good practice in relation to official statistics includes ensuring their accessibility.”
Section 8 deals with monitoring and reporting of official statistics: -
“(1) The Board is to monitor the production and publication of official statistics.
(2) The Board may report any concerns it has about—
(a) the quality of any official statistics,
(b) good practice in relation to any official statistics, or
(c) the comprehensiveness of any official statistics, to the person responsible for those statistics.
(3) The Board may publish its findings or any report under this section.”
Section 9 provides: -
“(1) The Board is to—
(a) develop and maintain definitions, methodologies, classifications and standards for official statistics, and
(b) promote their use in relation to official statistics.
(2) The Board may in particular for the purposes of this section give guidance and advice to persons responsible for official statistics.
(3) The Board may publish guidance and advice given under this section.”
Section 10 requires the UKSA to prepare adopt and publish a Code of Practice for Statistics.
At the request of an “appropriate authority” the UKSA must assess and determine whether any official statistics comply with the Code of Practice under s.10 (s.12(1)). The UKSA must designate statistics which do so comply as “National Statistics” (s.12(2)). The UKSA must publish the results of its assessment (s.12(6)). An “appropriate authority” includes a Minister of the Crown or, in relation to statistics produced by the UKSA, the National Statistician (s.12(7)).
The effect of designating a statistic as a National Statistic is that those producing it become subject to a duty under s.13(1) to ensure that it continues to comply with the Code under s.10. Failure to comply with that duty may result in the designation being cancelled under s.14.
Section 14(1) requires the UKSA to assess and determine whether designated National Statistics continue to comply with the Code of Practice. If the Authority determines that a National Statistic does so comply it must confirm the designation; otherwise it must cancel that designation (s.14(2)). The results of the assessment must be published (s.14(5)).
The Head of Assessment is the principal adviser to the UKSA on the exercise of its functions under ss.12 and 14 (see s.33). By s.31(3) the National Statistician may not exercise the functions of determining whether to adopt a Code of Practice under s.10 or to determine under ss.12 or 14 whether any official statistics comply with that Code. There is therefore a separation of function within the UKSA between the National Statistician and the Head of Assessment, which extends also to employees involved in the production of statistics (see also s.34).
Section 15 requires the UKSA to prepare and publish a programme for assessments under ss.12 and 14. The Authority must also prepare and publish a statement of the principles to which it will have regard, and the procedures it will adopt, in carrying out its functions under ss.12 and 14.
UKSA must publish at least once a year a list of the statistics currently designated as National Statistics. The court was shown the current list. It is extensive and, as is to be expected, covers a huge range of subjects.
Section 20 empowers the UKSA to produce and publish statistics relating to any matter concerning the UK or any part of it, and to publish information or give advice on those statistics.
Section 21 deals with the RPI: -
“(1) The Board must under section 20—
(a) compile and maintain the retail prices index, and
(b) publish it every month.
(2) Before making any change to the coverage or the basic calculation of the retail prices index, the Board must consult the Bank of England as to whether the change constitutes a fundamental change in the index which would be materially detrimental to the interests of the holders of relevant index-linked gilt-edged securities.
(3) If the Bank of England considers that the change constitutes a fundamental change in the index which would be materially detrimental to the interests of the holders of relevant index-linked gilt-edged securities, the Board may not make the change without the consent of the Chancellor of the Exchequer.
(4) In this section —
“index-linked gilt-edged securities” means securities issued under section 12 of the National Loans Act 1968 (c. 13) the amount of the payments under which is determined wholly or partly by reference to the retail prices index;
“relevant index-linked gilt-edged securities” means index-linked gilt-edged securities issued before the commencement of this section subject to a prospectus containing provision relating to early redemption in the event of a change to the retail prices index;
“retail prices index” means the United Kingdom General Index of Retail Prices.”
I note the following points in relation to s.21: -
The SRSA does not contain any definition of the RPI other than that given in s.21(4). The statute does not go any further. It does not, for example, identify any intrinsic or immutable characteristics of that index;
The UKSA is under a duty to continue to compile, maintain and publish that index. That duty continues unless and until s.21 is repealed or amended;
It is common ground between the parties that the UKSA has the power to make changes, including “fundamental changes”, to “the coverage and the basic calculation of the retail prices index”;
Before making any change to that coverage or basic calculation the UKSA must consult the Bank of England as to whether it would be a fundamental change materially detrimental to the interests of the holders of index-linked gilts containing a provision for early redemption in the event of a change to the RPI (i.e. an old-style gilt);
If the Bank of England so considers, the UKSA may not make the change without the consent of the Chancellor;
The criterion which the Bank of England has to apply is the same criterion as that which triggers the right of holders of old-style gilts to exercise the early redemption clause;
The object of ss.21(2) to (4) is simply to provide the Chancellor with an opportunity to veto a change in the RPI which could trigger such early redemptions. That provision does not confer any right to protection upon stockholders;
This power of veto was introduced in 2007 after it had been decided to replace early redemption clauses in index-linked gilts with cessation clauses (see [9] and [10] above). The veto has no relevance to index-linked gilts issued from 2002 or for any wider purpose.
Section 26(1) gives the UKSA a general ancillary power to do anything it considers necessary or expedient for the purpose of, or in connection with, the exercise of its functions. That would include amending the RPI, an index which has evolved and been amended throughout its past.
The Code of Practice for Statistics
Section 19 of the SRSA 2007 allowed the UKSA to continue to use the National Statistics Code of Practice in existence when that section came into force until the Authority adopted its first code under s.10. It is reasonable to assume that that Code was available to Parliament and informed its enactment of ss. 10 and 19.
In January 2009 the UKSA issued its first Code of Practice for Official Statistics (“the 2009 Code”), described as a “cornerstone of the new statutory framework”. The Authority intended that designation as a National Statistic should be seen as “a stamp of assurance that the statistics have been produced and explained to high standards and that they serve the public good”.
The 2009 Code was said to be consistent with the United Nations Fundamental Principles of Official Statistics which at the time included the following: -
“Principle 1. Official statistics provide an indispensable element in the information system of a democratic society, serving the Government, the economy and the public with data about the economic, demographic, social and environmental situation. To this end, official statistics that meet the test of practical utility are to be compiled and made available on an impartial basis by official statistical agencies to honour citizens’ entitlement to public information.
Principle 2. To retain trust in official statistics, the statistical agencies need to decide according to strictly professional considerations, including scientific principles and professional ethics, on the methods and procedures for the collection, processing, storage and presentation of statistical data.
Principle 3. To facilitate a correct interpretation of the data, the statistical agencies are to present information according to scientific standards on the sources, methods and procedures of the statistics.”
The 2009 Code contained eight “principles” and associated “practices” which taken together were intended to ensure that the range of official statistics meets the needs of users and are produced to high standards. Principle 1 stated: -
“Principle 1: Meeting user needs
The production, management and dissemination of official statistics should meet the requirements of informed decision-making by government, public services, business, researchers and the public”.
One relevant “practice” stated: -
“Investigate and document the needs of users of official statistics, the use made of existing statistics and the types of decision they inform”
Principle 2 required official statistics to be managed impartially and objectively.
Under the heading “Integrity”, principle 3 required the public interest to prevail over organisational, political or personal interests at all stages in the production, management and dissemination of official statistics.
Principle 4 required inter alia statistical methods to be consistent with scientific principles and internationally recognised best practices. The relevant “practices” included ensuring that official statistics are produced to a level of quality that meets users’ needs and seeking to achieve “continuous improvement in statistical processes”.
The edition of the Code of Practice published in February 2018 was current at the time of the decision under challenge. The principles in the 2009 Code regarding integrity, impartiality and objectivity, sound methods, recognised standards, accuracy and relevance to users continued to permeate the 2018 Code.
A comparison of the RPI and CPIH
In her consultation document on the RPI published on 14 November 2012 the National Statistician explained that both the RPI and CPI are designed to measure changes over time in the prices of goods and services consumed by households. Each index relates to a very large number of such items, referred to as a “basket”. As prices change, so does the total cost of all items in that basket. The indexes measure these changes from month to month. Both the RPI and CPI measure average inflation in relation to average household expenditure.
Between 2015 and October 2021 Mr Athow was the Deputy National Statistician and Director General for Economic Statistics at the ONS. He explains that the RPI, like other indexes, has been the subject of many changes over the years as part of a process of improvement and evolution.
For example, between 1993 and 2013 the changes made to the RPI included: -
the community charge element was replaced by the council tax and foreign holidays were included for the first time (1993);
depreciation was included in housing costs and domestic holidays were added (1994);
hedonic methodology was imported from the CPI to estimate inflation on certain goods (2004);
changes were made to price collection for some clothing and to the measurement of mortgage interest rates (2010);
improvements were made to the treatment of seasonal items (2011);
changes were made to the assessment of car prices (2012);
changes were made to the collection of data on private housing rents (2013).
Changes have also been made to the CPI from time to time.
In R (FDA) v Secretary of State for Work and Pensions [2013] 1 WLR 444 Lord Neuberger MR (as he then was) gave a helpful summary of the RPI and CPI and of certain differences between them ([9] – [21]). He said that they represent the two main measures of consumer price inflation. Both measure the average change in price of a fixed basket of goods and services over time. In 2010, the CPI was seen by the ONS as the main measure of inflation for macro-economic purposes and the RPI was “the most longstanding general purpose measure of inflation in [the UK]”. At [13] Lord Neuberger said: -
“There are a number of similarities in the ways in which RPI and CPI are determined. They are both calculated by reference to representative goods and services. Each year the ONS identifies a “shopping basket” of around 700 representative goods and services on which consumers typically spend their money. The items will be changed each year so as to ensure that they reflect changes in the pattern of consumer spending. It is the movement in the price of these goods which is used to measure the relevant price changes. Prices are obtained from many outlets, and an overall inflation rate is worked out by a process of, first, aggregating particular items into defined categories of products and calculating an inflation rate within each category, and then by weighting those categories and the inflation rates within them so as to produce a single overall figure for inflation.”
Mr. Athow explains that at the lowest level for aggregating items, weights are not available from household expenditure surveys. Instead, an unweighted formula is used to combine such items so as to arrive at “elementary aggregate indices” which can then be aggregated using household expenditure weightings as part of the process of building up an overall price index (WS para. 20). To calculate the elementary aggregates the RPI uses the “Carli” formula which employs arithmetic averaging. Both the CPI and CPIH primarily use the “Jevons” formula which relies on geometric averaging. The effect of the mathematical properties of the Carli formula is that the price inflation it calculates cannot be lower than that given by the Jevons formula for the same elementary aggregate. This difference between the RPI and the CPI (or CPIH) is referred to as the “formula effect” (WS para. 27). A guide produced by the United Nations in 2009 referred to the upwards bias of the Carli formula relative to trends in the average price of items and advised that it should be avoided (WS para. 30).
From the late 1970s onwards countries started to move away from the Carli formula to the Jevons formula. By 2012 the UK was the only one of 28 European countries which still used the Carli formula for computing a national price index (WS paras. 25-31).
Between 2010 and 2018 the formula effect added on average 0.7 of a percentage point to the RPI compared to the CPI (WS para. 32). In 2010 a routine improvement in the collection of price data for clothing resulted in the formula effect nearly doubling from around 0.5 to around 0.9 of a percentage point (WS para. 35). It is said that this artificial increase in the measure of inflation has resulted over many years in higher expenditure for those paying interest on debt or costs linked to the RPI and a windfall gain for those entitled to receive income index-linked to the RPI.
In 2017 the Office for Budget Responsibility estimated that the long run effect of the overall differences between the RPI and the CPI, referred to as “the wedge”, amounted to a whole percentage point.
Another significant shortcoming of the RPI is its treatment of housing costs. Mr. Athow has explained why the CPIH is superior because, in particular, it is based on rental equivalence rather than the cost of funding the purchase of a house and measures reflecting asset value (WS paras. 81 to 93).
Mr. Athow summarises a number of other shortcomings in the RPI which are addressed by the CPIH (WS paras. 94 to 109). In summary: -
Population coverage
The RPI excludes the top 4% of households by income and some of the poorest pensioner households, amounting to 12% of households in total. Those households are included in the CPIH.
Weights for household expenditure
The data used for the RPI tends to involve small sample sizes and only relates to private households. It excludes institutional households such as care homes and halls of residence. The data used for the CPIH is more comprehensive and accurate.
Foreign visitors and domestic expenditure
The CPIH adopts a consistent approach whereas the RPI does not.
Price Collection
The CPIH addresses price volatility within a month better than the RPI.
Classification
The classification of goods and services by the RPI is inferior to that of the CPIH.
Addressing the flaws in the RPI
This section should be read together with the Introduction to this judgment.
In October 2012 the National Statistician launched a consultation on options for improving the RPI, following a programme of work to better understand the use of the Carli formula and the formula effect. In addition to “no change”, three options were put forward to deal with the formula effect. They each involved abandoning the use of the Carli formula, whether for clothing or more generally, and alternatives for replacing it with inter alia the Jevons formula. That consultation only proposed changes in the formulae for “aggregation” and not other matters, such as the coverage of, and weightings used in, the RPI.
On 10 January 2013 the National Statistician announced her recommendations which had been accepted by the UKSA. First, the use of the Carli formula in the RPI did not meet international standards. Accordingly, an additional RPI-based index using Jevons’ geometric formula would be published from March 2013, known as the RPIJ. In relation to the RPI the announcement said: -
“In developing her recommendations the National Statistician also noted that there is significant value to users in maintaining the continuity of the existing RPI’s long time series without major change, so that it may continue to be used for long-term indexation and for index-linked gilts and bonds in accordance with user expectations.
Therefore, while the arithmetic formulation would not be chosen were ONS constructing a new price index, the National Statistician recommended that the formulae used at the elementary aggregate level in the RPI should remain unchanged.”
In February 2013 the ONS published a Response to the consultation exercise. It stated that it is not uncommon for the ONS to adjust the methodology used to calculate an index as improved methods become available, but the ONS “also recognises the importance of a consistent approach to an index that has many contractual, long-standing and long term uses” (p.13).
In March 2013 the UKSA published an assessment report on the RPI under s.14 of the SRSA 2007. The assessment applied the 2009 Code of Practice, in particular principles 1, 3 and 4. In relation to principle 1, meeting “the requirements of informed decision-making by government, public services, researchers and the public” paragraph 3.2 said this on the decision to continue producing the RPI with the Carli formula: -
“This decision is intended to meet the needs of certain users for continuity in the existing RPI – for example, in the index-linked gilts and bonds market, in private sector business contracts, and in the indexation of private pension payments.”
Plainly, that decision did not purport to take into account the needs of all users where monetary sums are linked to a measure of inflation. Indeed, the Consumer Prices Advisory Committee observed that competing user needs would be “better balanced” by replacing the Carli formula with Jevons.
Paragraph 1.2.2 of the Assessment stated that the effect of the decision to retain the basic formulation of the RPI for the benefit of certain users, was that the index was inconsistent with the 2009 Code. In particular, in relation to principle 4 “Sound methods and assured quality”, para. 3.4 stated: -
“Compliance with the Code requires that statistical methods be consistent with internationally recognised best practices. As part of its investigations of the ‘formula effect’, ONS researched approaches used by other national statistical institutes. In line with the International Labour Organisation’s Resolution concerning consumer price indices, all of the statistical institutes reviewed choose between two formulae at the elementary aggregate level, the Dutot and the Jevons. Only the UK was found to use the Carli. The National Statistician reported that ‘the ONS research programme found that use of the arithmetic formulation (known as the ‘Carli’ index formula) in the RPI is the primary source of the formula effect difference between the RPI and the CPI, and that this formulation does not meet current international standards’. On the basis of the evidence presented by the National Statistician, the Assessment team therefore considers that the RPI does not comply with the Code.”
Accordingly, the UKSA cancelled the designation of the RPI as a National Statistic. But paragraph 2.7 of the Assessment stated that for the foreseeable future ONS would continue to produce and publish the RPI because of the user needs referred to in [83] above.
The report of the Johnson Review was published in January 2015. The Review recommended that Government and regulators should work towards ending the use of the RPI as soon as practicable. Where they decided to keep using it they should give their reasons for doing so. The introduction of the RPIJ shared some of the shortcomings of the RPI and had caused confusion amongst users and so it should be discontinued (in fact it was discontinued in March 2017). The Review also recommended:-
“The logic of the National Statistician’s recent decisions is that the RPI should be considered a legacy measure to be used only where contractually required. No further changes should be made to the RPI. If a change is made to the CPI and CPIH that would affect the RPI, the production of the indices should be split to retain the best practice of the CPI and CPIH and the constancy of the RPI. Over the long term the Authority should look to phase out production of the RPI in consultation with users, amending the law (the Statistics and Registration Service Act 2007) as necessary.”
In June 2015 the UKSA established two independent advisory panels on consumer price statistics. It also carried out public consultation on the Johnson Review. The Authority stated that the use of the Carli formula resulted in “a consistent upward-bias” (p.4). There were “basic statistical flaws” in the RPI, but given its use in commercial contracts and index-linked gilts, the UKSA did not propose to discontinue the RPI, which was described as a “legacy index” (p.12).
On 9 March 2016 the National Statistician wrote to the Chair of the UKSA to set out the approach he intended to take. With regard to the RPI he said: -
“Third, users have sought clarification on the future of the Retail Prices Index (RPI). Put simply, I believe that the RPI is not a good measure of inflation and does not realistically have the potential to become one. I strongly discourage the use of RPI for (sic) as a measure of inflation as there are far superior alternatives. Nonetheless, RPI is still used for a number of legacy purposes and its production is mandated by legislation. My intention is that from the start of 2017, ONS would publish the minimum of RPI-related data necessary to ensure the critical and essential needs of existing users are met. This would consist of the aggregates RPI and RPIX. Component indices of RPI (for example, the RPI indices for food, clothing etc) would also continue. For the avoidance of doubt, RPIJ would no longer be published.
The RPI would continue to be maintained through routine changes. This covers all changes required to continue production of a consistent, fit for purpose RPI (for example the annual update of the basket and weights, computer systems upgrades and improvements to data validation and quality assurance methods). With due consideration to the requirements of the Statistics and Registration Service Act 2007, ONS would only consider making methodological changes to the RPI if to not do so would inhibit the improvement of CPIH and the Consumer Prices Index.”
On 8 March 2018, not long before the House of Lords Economic Affairs Committee began its inquiry, the ONS published a document entitled “Shortcomings of the Retail Price Index as a measure of inflation”. It stated that ongoing work by the ONS and others had strengthened the case against the RPI. The document explained the many flaws in the index in addition to the use of the Carli formula.
As we have seen, the report of the House of Lords Committee introduced into the discussion (a) the interests of all those affected by the use of the RPI, and not simply those holding index-linked investments and (b) the need for the flaws in the index to be overcome by methodological improvements.
The UKSA Board met on 14 February 2019. They received oral advice from the National Statistician which was reproduced as a document dated 26 February 2019. He advised that the CPI and CPIH had been designed to be statistically robust and to satisfy the obligation in s.7 of the SRSA 2007 to promote and safeguard statistics for the public good (para. 6). The treatment of the RPI in s.21 of the Act was problematic from a statistical point of view because a “legislative requirement that treats changes in one direction differently from another implies bias”. It is common ground that he could only have been referring to the power in s.21(3) to veto fundamental changes to the index which would harm the interests of the holders of gilts with early redemption clauses.
The National Statistician advised the Board on four options. He advised against the first, which was to maintain the current position. It was no longer possible to maintain the RPI with only “routine maintenance”. Six years after the RPI had ceased to have National Statistic status, it was still being used widely. Something more was needed (paras. 12-14).
He also recommended against the second option, namely to “fix” the clothing element of the RPI. There were other shortcomings in the RPI, some of them serious. It would be arbitrary to tackle just the clothing element by substituting the Jevons formula for Carli. Furthermore, there were solutions for all of the shortcomings (paras. 15-19).
The third option was to correct all the shortcomings of the RPI (paras. 20 to 26). There were different ways of conceptualising this approach but essentially it would mean using the methods and data of one of the alternative indexes which are statistically robust. He identified two statistical issues with this approach. The second was the bias concern previously mentioned (see [91] above). Paragraph 24 stated:-
“Secondly, under this approach section 21 would still apply and if we permanently aligned RPI and CPIH, all changes to CPIH would need to go through the process set down in legislation. As noted above, the legislation treats changes that increase RPI differently from changes that reduce it. This is a biased methodology. This problem could be avoided if we allowed RPI and CPIH to diverge in future, but then many of the benefits of alignment would be lost and the statistical quality of RPI could deteriorate once again”
Nevertheless, he said that using s.21 to align the RPI with the CPIH would be statistically more robust than simply fixing the clothing element.
The National Statistician considered the fourth option, ending the publication of the RPI, to be the strongest from a statistical viewpoint, because of the removal of the bias in s.21 (paras. 27-28). The UKSA would be left with only statistically robust measures of inflation. Although this option would depend upon legislation to repeal s.21, it was the one recommended by the National Statistician. He concluded by saying that he could also support the third option, to correct all the shortcomings of the RPI by aligning its “coverage and construction” with the CPIH.
The minutes of the Board’s meeting record the Chair advising members that, although they were aware of the wider implications of changing the RPI, the Board should address the issues “on the basis of statistical considerations only” (para 3.2), as the National Statistician had done (para. 4.3).
The Board agreed to “propose two parallel actions” (para 4.8): -
“i. that that (sic) RPI publication is stopped at a point in the future. This would require primary legislation to amend the Statistics and Registration Service Act 2007, and is therefore likely to be a longer-term solution; and
ii. in the interim, to correct the shortcomings of RPI by bringing the methods of CPIH into it. If adopted, this would mean that at least initially the RPI would be the same as CPIH. There would then be an ongoing question whether future changes to the CPIH should also be reflected in changes to the RPI.”
The Board decided that “this approach” should be set out in letters to the Bank of England and the Chancellor, plainly envisaging that the procedure under ss.21(2) and (3) would be invoked in relation to the third option.
I see no merit in the claimants’ suggestion that because the Board used the word “propose” (and elsewhere the UKSA or ONS used the word “proposal”) it did not take a decision at its meeting on 14 February 2019. The Board plainly did decide to approve two options which should be advanced in parallel, one a future but permanent change and the other an interim solution. They used the word “propose” because both options depended upon additional decision-making by others, the first under s.21 of the SRSA 2007 and the other requiring the enactment of new legislation by Parliament.
There were real differences between the third and fourth options. First, the third option would not result in the disappearance of the RPI, and its replacement by the CPIH, just as if s.21 had been repealed. The RPI would remain, but its methodology and coverage would initially be aligned with the CPIH. Second, although initially there would be no statistical difference between them, they would remain in substance two separate indexes. Each subsequent change to the CPIH would give rise to a separate question as to whether the RPI should also be changed which, until 2030, would also require the s.21 procedure to be followed. Third, the alignment of the RPI with the CPIH (subject to approval under s.21) would not overcome the statistical bias in s.21. Fourth, the third option was seen as an interim solution, pending any decision to repeal s.21.
The approach set out in the National Statistician’s advice to the Board’s meeting and reflected in its minutes, was faithfully conveyed by the Chair of the UKSA in his letter to the Chancellor dated 4 March 2019. Because there has been a good deal of misunderstanding in relation to this letter, it is necessary to quote virtually all of it:-
“This letter sets out the UK Statistics Authority’s proposals for the future of the UK’s consumer price statistics. It describes the current position, a changing context, and – drawing on the advice of the National Statistician – proposals for change. In formulating the substance of our approach, the UK Statistics Authority may only take statistical matters into account. We recognise that our proposals would have substantial wider effects but these are not matters that can bear on our decisions.
The question whether to make changes to the RPI to improve it was the subject of consultation in 2012. The decision made by the then National Statistician, and one widely supported in the consultation at the time, was to leave the RPI unchanged. This decision gave rise in turn to the conclusion that the RPI should be treated as a legacy measure, with no future substantive changes to its construction and methods. That position was endorsed by an independent review of consumer prices led by Paul Johnson and confirmed in the UK Statistics Authority’s response to the 2015 consultation on the Johnson review. In the period since, the Office for National Statistics (ONS) has developed alternative measures of inflation, and the Authority has urged users to move away from the RPI.
Since the Authority’s decision, in 2013, to maintain the RPI as a legacy index, several factors have changed. In 2017, the CPIH was designated as a National Statistic and the Household Cost Indices were first published. In 2018, ONS published its analysis of shortcomings of the RPI, summarising the various issues. In January the Lords Economic Affairs Committee’s report made a clear call for the UK Statistics Authority to reconsider its position on RPI. And earlier this month, the National Statistician’s Advisory Panel on Consumer Prices concluded that the current position is unsatisfactory and urged us to take action on the RPI.
Some six years since the RPI lost its National Statistic status, it continues in widespread use. We are in a position where we and the ONS, who are the producers of the UK’s consumer price statistics, are clear that the RPI is not a good measure of inflation. It is however the only statistic that we are legally obliged to produce and which we cannot change without consulting the Bank of England and potentially you. Its production, and the legislation around it, sits uncomfortably with our legal obligation to promote and safeguard the quality of official statistics. The legislation is particularly problematic from a statistical perspective as it treats changes that increase the RPI differently from changes that reduce it.
Our recommendation, in line with our long-standing position, is that publication of the RPI should cease. Better alternatives exist and to announce the end of the RPI for an appropriate future date would give the Government, the markets, businesses and others time to agree on alternative arrangements.
Recognising that the abolition of the RPI will require primary legislation, that the legislative route would take time, and that there would be substantial implementation issues that the Government would need to consider, the Authority proposes a second course of action to be undertaken in parallel
Until 2030, when the last relevant index-linked gilts mature, we are constrained by section 21 of the Statistics and Registration Service Act 2007. This says that if a change is proposed to the RPI that is both fundamental and materially detrimental to the holders of those index-linked gilts, it could only be made with your consent. The Act requires that the Bank of England should determine whether any proposed change meets the test for consulting you. I therefore wrote to the Governor to describe the change we propose. This would address the shortcomings of the RPI by adopting the methods of the accredited CPIH measure. The effect, at least initially, would be to turn the RPI into CPIH by another name.
The Bank of England has confirmed that such changes would be both fundamental and materially detrimental. I enclose a copy of the Bank of England’s advice to me.
We recognise that this would be a substantial change to the RPI, with its own disadvantages. We would be favouring a particular index to replace the current RPI, rather than allowing individuals and firms to choose the index with the most appropriate statistical properties for their circumstances. In addition, the requirements set down in section 21 would still apply. So, there would be a continuing question whether future changes to CPIH should also be reflected in the RPI, with the potential for the indices again to move apart in future. However, there are no options for the future of the RPI that are without challenges, and we believe the present situation is unsustainable.
I would therefore be grateful to know whether you are willing to give your consent under section 21 of the Statistics and Registration Service Act 2007 to make the changes to the RPI described here. We recognise that there are a number of wider issues you will need to consider, and should you agree, we would be happy to discuss with the Treasury the timing and mechanics of the change.
It would of course be open to you to decide to accept or reject separately each of abolition and the proposed changes to the RPI. But our clear recommendation is that both should be pursued.” (italics added)
Mr. Herberg QC on behalf of the claimants repeatedly emphasised during his submissions the sentence I have italicised in that quotation. He suggested that this showed that: -
The RPI decision was an attempt to circumvent the unlikelihood of s.21 being repealed and to achieve the same outcome by a different means. It was a “workaround”;
In reality, the RPI decision amounted to replacing the RPI with the CPIH and rebadging the CPIH as the RPI.
These suggestions had overtones of an ulterior or improper purpose, but Mr. Herberg made it plain that he was not alleging that. In any event, both suggestions are completely untenable on any fair reading of the letter of 4 March 2019. That letter must be read as a whole and in the context of the National Statistician’s advice and the Board’s minutes. The sentence emphasised by the claimants must be read properly in its context as set out in [94]-[100] above and not wrenched away from it.
The fourth paragraph of the quotation from the letter of 4 March 2019 ([101] above) very fairly set out the problems caused for UKSA by s.21, including the tension created between its legal obligations under s.7 and the “bias” effect of s.21.
It will also be noted that although the UKSA recognised that its proposals would have “substantial” wider effects, it considered that those matters could not affect its decision-making: -
“the UK Statistics Authority may only take statistical matters into account”
In his letter of 30 April 2019 the Chancellor outlined some concerns about the UKSA’s proposals. He accepted that “the RPI is a flawed measure of inflation” and that “its publication, and its continued widespread use after flaws have come to light, potentially affects the credibility of the wider UK statistical system”. He accepted that the flaws needed to be addressed and understood the arguments in favour of aligning the RPI with the CPIH. But the Chancellor was concerned that a reduction in the RPI’s measured rate of inflation would impact on those who receive income linked to the index. In addition, the UKSA had not consulted publicly before making its proposals, which would reverse the decision of the National Statistician in 2012 to maintain the RPI in its then form.
The response of the Chair of the UKSA on 1 May 2019 included the following:-
“We of course understand that our proposals would have major consequences for a broad range of people and institutions. However as I said in my earlier letter the legislation allows the UK Statistics Authority only to consider the statistical aspects of any decision, rather than the wider public policy issues that will be your concern. The legislation underpinning the RPI is clear that it is for the Authority, advised by the National Statistician, to make recommendations based solely on statistical quality, and for Government to consent to them or not.”
The Chancellor received briefings from his officials. They expressed concern that there had not been any prior consultation by UKSA on the principle of the RPI decision and that the Treasury lacked a full picture of how RPI is used across public and private sectors. But officials also advised that if the Chancellor were to reject the UKSA’s proposals, there would be pressure on Government from those who lose out because of the flaws in the RPI.
In their advice to the Chancellor on 1 August 2019, officials identified one crucial factor as being the ability of the UKSA to act unilaterally to align the RPI with CPIH from 2030. Having discussed impacts on winners and losers through either aligning the RPI with the CPIH or keeping it unaltered, officials advised:-
Officials’ view is that there is merit in aligning RPI with CPIH as the long-term solution to fix the flaws in RPI. It would solve the difficulties involved in moving private sector contracts from a flawed measure of inflation, by moving them automatically in line with an accredited National Statistic. Similarly, by adjusting both assets and liabilities linked to RPI automatically and at the same time, it would help to maintain a unified, deep and liquid market for the government’s index-linked debt, without the fragmentation from building a CPI or CPIH linked market alongside that for RPI. Finally, alignment would address the distributional inequities from the use of a flawed inflation measure.
If you wish to pursue this option, the right thing to do would be to consult on when to make the change. UKSA have not set out their preferred timeframe nor consulted publicly with those that could be affected by the change. A joint consultation with UKSA on when to align RPI with CPIH would allow for further investigation of the impacts, and provide government with a basis for considering the appropriate notice period before alignment happens.”
The briefing concluded by saying that the judgments on the issues involved were finely balanced.
On 4 September 2019 the Chancellor responded formally to the UKSA’s request for his consent under s.21 and announced a joint consultation on the same day (see [21]-[22] above).
The joint consultation document issued on 11 March 2020 explained that the Chancellor’s role under s.21 of the SRSA 2007 arose from certain index-linked gilts. Because his decision on consent could only consider factors related to index-linked gilts, the Chancellor sought responses on the potential impact of the UKSA’s proposal on the holders of such gilts and potential broader impacts on the index-linked gilt market (paras. 5-6).
Paragraph 7 of the consultation document continued: -
“The Authority and the government are also mindful that they do not have full sight of the use of the RPI in the economy and financial contracts, and that the change to the RPI could have unintended and diverse impacts. The Authority is also aware that its intended method of implementing its proposal to address the shortcomings of the RPI will make publishing sub-indices of the RPI no longer feasible. These factors are unlikely to be relevant to any decision that the Authority is minded to make as regards addressing the shortcomings of the RPI or that may fall to the Chancellor under legislation. But this consultation welcomes evidence on the use of the RPI and its sub-indices more widely to inform future policy decisions. Section 6 sets out how and why the Authority and government are seeking evidence on these broader issues, which are likely to be outside of scope of the potential decisions by the Authority and the Chancellor referred to in other sections of this document”
Paragraph 40 stated: -
“Beyond gilts, the RPI is used widely in the economy. The impacts on these wider uses from the Authority’s proposal being implemented are likely to fall outside the factors the Chancellor can consider in his decision under section 21 of the Act. For instance, the RPI is used to uprate some taxes and benefits; to determine changes in rail fares, reflecting industry costs; and to determine the rate of interest on student loans. In the private sector, the RPI is used in some wage agreements; to uprate certain pension payments, particularly defined benefit pensions; in calculating rent increases for some leasehold properties; and in financial markets. The government welcomes information on the potential wider impacts from the Authority’s proposal.”
In para. 38 of the consultation document, the UKSA said that “it remains minded, subject to the requirements of section 21, to implement” the RPI decision, and that there were two key elements which had yet to be established and were now being consulted upon. They were firstly, the technical approach for bringing the methods and data sources of the CPIH into the RPI and secondly, a specific date at which the Authority’s proposal would be implemented and the impacts of that for the Chancellor’s decision under s.21.
Mr. Herberg relied upon the use of the words “minded to” as indicating that no decision had yet been taken on the principle of aligning the RPI with the CPIH. I disagree. The document gave the straightforward explanation that the UKSA’s intention to alter the RPI remained subject to the s.21 consent procedure if implemented before 2030. Moreover, the Authority could not align the RPI to the CPIH before February 2021 because of the need to make changes to the UKSA’s systems for producing the RPI. In addition, the Chancellor had already said that he would not give his consent under s.21 to the proposed change before February 2025 (see [21] above). But if the Chancellor decided to give his consent, there is no reason to think that the UKSA would not have implemented the change as soon as practicable. Once again, the claimants’ suggestion that no decision had been taken by the UKSA to align the RPI to the CPIH before the consultation in 2020 is wholly unrealistic.
Section 5 of the consultation document dealt with the timing for implementing the UKSA’s proposal during the window between 2025 and 2030. Paragraph 25 identified considerations relevant to the Chancellor’s decision under s.21. These were elaborated at paras. 58 to 60. They included the effect on the public finances of the Government having to redeem old-style gilts, the potential reduction in the cost of Government debt through a reduction in the RPI, the effect on financial institutions holding RPI-linked gilts to hedge RPI-linked liabilities and the impacts on the gilt market. The Chancellor expressly sought responses on those matters.
The joint Response to the Consultation was published on 25 November 2020. Paragraph 43 referred to the number of respondents who had said that the consultation was too narrow; it should have focused on whether the change to the RPI should be made at all.
The Response noted that the consultation had yielded a range of information on the index-linked gilt market. The effect of the change to the RPI would vary according to the nature of the arrangements made by different financial institutions. For example, at one end of the spectrum some schemes with perfectly hedged provision might be no worse off at all (para. 69). At all events, the vast majority of index-linked gilt investors indicated a strong preference for the proposal to be implemented as late as possible, that is in 2030. This was to allow them as much time as possible to adjust to the change and to minimise any potential negative impacts (para.75). The Chancellor explained why he would not give his consent under s.21 to the RPI being changed before 2030 (paras. 77-80).
Lastly, the Response document summarised information gathered on the broader effects of the change to the RPI (paras. 95-100). Put shortly, the information identified winners and losers from aligning the RPI to the CPIH.
Effects on legacy users of the change to the RPI
In this section I will only summarise effects described in more detail in the evidence.
Mr. Groom explains that as awareness of the risks associated with defined benefit pension schemes increased from the mid-2000s, funds moved away from investing in equities for the purposes of matching their benefit liabilities to investing in assets the value of which is said to move in line with the value of all or some of those obligations. This is referred to as liability driven investment. It is seen as being consistent with the duty of pension fund trustees to invest in accordance with the “prudent person” rule. The practice is also encouraged in guidance issued by the Pensions Regulator. Assets managed by institutional investors, including pensions schemes, under the liability driven approach have tripled from £400 billion in 2011 to £1.2 trillion in 2018. Over £1 trillion of private sector defined benefit pension liabilities were hedged in that way in 2018. Generally, liability driven investment has relied upon index-linked gilts.
Mr. Groom explains that in 2011 HM Treasury and the Debt Management Office stated that the issuance of CPI-linked gilts would be unlikely to be cost-effective in the near term and would involve risks. He says that the supply of other CPI-linked instruments is insufficient to meet demand and the market is illiquid.
Mr. Groom refers to estimates that the RPI decision has caused the value of index-linked bonds held by defined benefit pension schemes to decline by £60 billion or 13%. It is also estimated that the decision has caused holders of all index-linked gilts currently issued with a maturity after 2030 to suffer a loss of £90 billion - £100 billion (reflecting a reduction in all coupon and principal payments from 2030). This has reduced the value of the assets held by a pension fund and hence its funding ratio, i.e. the ratio of those assets to the liabilities of the fund.
These effects have occurred because, it is said, the RPI decision has caused the market to reprice existing RPI-linked instruments as if they will be linked to the CPIH from 2030, which involves a reduction of about one percentage point in that future income stream.
The first claimant has 82,000 members who are entitled to have their pensions increased in line with the RPI. It is estimated that the effect of the RPI decision has been to reduce the present value of their pensions by £2.8 billion.
As at September 2019 the first claimant held over £20 billion of index-linked gilts, inflation swaps and RPI-linked corporate bonds. It is estimated that the effect of the RPI decision has been to reduce the value of those assets by £3.7 billion. Relative to the fall in the present value of RPI linked pensions, there is a deficit of about £1 billion.
Mr. Oakley and Mr. Garner present a similar picture in relation to the second, third, fourth and fifth claimants. The point is made that even where pension benefits are linked to the CPI rather than the RPI, pension funds have had to rely mainly on RPI-linked gilts to match, or hedge against, inflation-protected benefit liabilities.
Mr. Oakley gives evidence specifically on the disproportionately greater impacts which the RPI decision has on the present value of pensions for younger and female members of the scheme.
Ground 1
The claimants submit that no explicit consideration was given by the UKSA or by the Chancellor to whether the RPI decision was intra vires s.21(1). They appear to have assumed that it was. This is said to be all the more surprising given their recognition that cessation of the publication of the RPI depended upon that provision being repealed. I do not attach any significance to these points, essentially for the reasons given by Laws J (as he then was) in R v Somerset County Council ex parte Fewings [1995] 1 All ER 513, 523. What matters is whether on the true interpretation of the SRSA 2007 the RPI decision fell within the Authority’s powers. That is an objective question of law for the court. Ultimately this was accepted by Mr. Herberg and rightly so (transcript day 2 p.141).
As to the submission that the RPI decision was simply an attempt to circumvent the need for s.21 to be repealed if publication of the RPI were to cease, I have already explained why that is untenable (see [94] – [104] above).
Accordingly, the claimants’ case is based upon two propositions: -
The RPI decision was, as a matter of substance, a decision to cease producing the index known as the RPI;
That approach was inconsistent with the UKSA’s duty under s.21(1) to produce and publish the RPI.
Mr. Herberg points to the language of s.21. Parliament has imposed a duty on the UKSA to compile, maintain and publish “the retail prices index” meaning the United Kingdom General Index of Retail Prices, and not simply “a retail prices index”. When the SRSA 2007 was enacted both the RPI and the CPI were in existence as two different indexes. Unlike the RPI, the CPI did not include housing costs. Parliament did not consider it appropriate to mandate the production of the CPI. The RPI is the only statistic which Parliament has required the UKSA to produce. Likewise, the provisions of the Pensions Act 2011 show that Parliament recognises a legal distinction between the RPI and the CPI.
Mr. Herberg accepts that the UKSA has a very broad discretion to amend the scope and methodology of the RPI from time to time. For example, it may make a “fundamental change” to the coverage or the basic calculation of the RPI (see s.21(2)). The UKSA can make a “fundamental change” to the RPI, as long as what results still has the legal characteristics of the RPI (transcript day 1, page 103). But it is not sufficient that, as the defendants say, the UKSA continues to publish an index which can rationally be described as a retail prices index (i.e. an index that measures the increase or decrease in UK retail prices).
The claimants submit that the wholesale adoption of the methods and sources of a different inflation index already in existence, the CPIH, will involve the substitution of that index for the RPI, or the replacement of the latter. The consequence is that the RPI will cease to be published. The fact that the term RPI will continue to be employed does not alter the analysis. In effect, the CPIH will be “rebadged” as the RPI, and the RPI referred to in s.21 of the SRSA 2007 will cease to exist.
In his reply Mr. Herberg submitted that the RPI is an index with a long history going back to 1947. The index which continues to be published under s.21 must be a descendent of that index.
Discussion
I agree with Sir James Eadie QC, who appeared on behalf of the defendants, that the claimants’ submissions have not really addressed the scope of the UKSA’s undoubted power to make “fundamental changes” to the RPI, or identified what are the legal characteristics of the RPI which may not be changed. As he put the question, what is it about the RPI which the claimants say sits above a fundamental change and is immutable?
The claimants sought to avoid that issue by repeating that the UKSA really wanted to see s.21 repealed and sought to circumvent the uncertainty of whether the legislation would be amended by “rebadging” the CPIH as the RPI. I have already rejected that argument.
Connected to that same argument, the claimants also sought to rely upon statements made on behalf of the UKSA that the RPI did not have the potential to become “a good measure of inflation”, as if to suggest that the index was considered to be incapable of reform, even by making fundamental changes, and so the UKSA’s intention must have been to replace it (see e.g. transcript day 1 p.108). Once again, the claimants sought to bring subjective intention into an issue which is agreed to be an objective question of law. The questions for the court still remain what power did the UKSA have under s.21 to alter the RPI and, as a matter of substance, did the RPI decision fall within it?
Equally unattractive is the argument (in para. 56 of the claimants’ skeleton) that the changes made by Parliament in the Pensions Act 2011 were premised on the natures of the RPI and the CPI being legally different. The claimants say that the “wedge” is “a well-recognised feature of the two inflationary indices” and a product of their different approaches to measuring inflation. They complain that it would be eliminated by the RPI decision. But that argument goes nowhere unless the wedge is a characteristic or differential which, in effect, the SRSA 2007 requires to continue. However, Mr. Herberg did not dispute the proposition that the RPI could lawfully be amended under s.21 by substituting the Jevons formula for the Carli formula in the RPI. That change alone would reduce the long run estimate of inflation produced by the RPI by 0.7 of a percentage point and eliminate most of the wedge.
In my judgment, there is nothing in the language of the SRSA 2007 to suggest that the statutory limits of the UKSA’s power to change the RPI are concerned with perpetuating “the wedge” or protecting any given level of return on investments linked to the RPI.
The Pensions Act 2011 does not assist the claimants. Section 51 of the Pensions Act 1995 requires pensions in payment under certain occupational schemes to be increased by a minimum percentage. Section 19(7) and (8) of the Pensions Act 2011 amended s.51 so that it refers to both the RPI and CPI, whereas the original enactment only referred to the RPI. But s.51 simply adjusts the requirement for a pension benefit to be increased by a minimum percentage where the rules of a scheme rely upon either the RPI or the CPI to increase the pension payable. Those indexes are not referred to for any other purpose. The legislation does not seek to define the characteristics of either index or any substantive difference between the two. Section 51 simply recognises that the rules of some pensions schemes rely upon the RPI and others rely upon the CPI.
In any event, the 2011 Act cannot be used to interpret the SRSA 2007 (R (Miller) v Secretary of State for Exiting the EU [2018] AC 61 at [113]). Subsequent legislation is not an interpretative tool for understanding the obligation in s.21 of the 2007 Act to maintain the RPI or the scope of the power to make fundamental changes to that index.
Having cleared the decks, I return to s.21 of the SRSA 2007. Parliament enacted a provision which referred to an index which, it must have known, had evolved and been altered substantially over many years. In part that has been because of changes in circumstance. But there have also been decisions to alter the coverage of the index (e.g. the inclusion of holidays and depreciation for housing) or the methodology used. Mr. Athow states that the RPI had previously been changed by incorporating data or methods used in other indexes. Up until the SRSA 2007, such changes would have been under the control of the Chancellor of the Exchequer. Under the Act they became the sole responsibility of the UKSA (subject only to the power of veto in s.21(3)).
Parliament did not find it necessary to confer or spell out an express power to change the RPI. Given the history and nature of the RPI as an index measuring consumer price inflation, it is obviously implicit in the duty in s.21 to compile and maintain that index that the UKSA is able to change it. The ancillary powers in s.26 also allow the UKSA to change the methodology or coverage of the RPI and other statistics it produces.
It is also common ground that by implication from s.21(2) the UKSA has the power to make fundamental changes in the coverage or the basic calculation of the RPI. This includes all aspects of methodology. The claimants accept that that power is not limited to fundamental changes affecting old-style gilts (transcript day 1 p.103).
The use of the word “fundamental” underscores how broad the Authority’s power to change the RPI is. Parliament could have chosen to place express constraints upon the scope of this broad power but it did not do so. Instead, it has entrusted the exercise of the power to change the RPI to the expert judgment of the UKSA, subject to judicial review applying Wednesbury principles, but with the degree of restraint, or enhanced appreciation, applicable to the technical decisions of an expert body (Mott v Environmental Agency [2016] 1 WLR 4338).
Parliament did not consider it necessary to provide in the legislation a definition setting out the main characteristics or purposes of the RPI, or matters which are not susceptible to change. Parliament did not require the retention of any specific features or qualities of the RPI as it existed at the time of enactment. The legislature did not mandate how the RPI should be compiled or what it should cover. Matters such as the contents of the basket of goods and services, weighting, aggregation and the use of formulae were left to the expert judgment of the UKSA.
I gratefully adopt the explanation of the RPI given by Lord Neuberger MR in FDA (see [72] above). The RPI, like the CPI (or for that matter the CPIH), is a measure of consumer price inflation. Each measures the average change in the cost of a defined basket of goods and services over time. Thus, the RPI and CPIH have essentially the same objectives. I do not accept that the CPI indexes belong to some legally different statistical family, as Mr. Herberg appeared to suggest. On the other hand, I accept the submission of Sir James Eadie QC that the RPI must be officially recognisable by being published by the UKSA as the United Kingdom General Index of Retail Prices.
Although the RPI is the only “official statistic” made subject to dedicated provisions in the SRSA 2007, including legal obligations, s.21 must be read harmoniously with s.7. Parliament has not provided for s.21 to be treated as an exception to s.7. The express language of s.7(1) and (3), which cross-refers to s.21, is to the contrary. The UKSA has the objective of promoting and safeguarding the production and publication of official statistics that serve the public good. The Board’s overarching objective in s.7(1) includes the duty in s.7(3) to promote and safeguard the quality of, and good practice in relation to, official statistics as well as their comprehensiveness. Although the designation of the RPI as a National Statistic was cancelled in 2013 because of a failure to comply with the 2009 Code, that does not alter the applicability of s.7 to the RPI under s.21. Those objectives and duties have remained relevant and they inform the scope and exercise of the power to make changes, or even fundamental changes, to the RPI.
There is no legal challenge to the UKSA’s identification of the flaws in the RPI. They include the use of the Carli formula, the handling of clothing prices, the treatment of housing costs, population coverage, weighting, price collection and coverage, and classification of goods and services. Not surprisingly, the claimants do not argue that the way in which each of these matters is dealt with in the RPI is an immutable characteristic of that index. Accordingly, each, and indeed all, of those matters can potentially be remedied by the UKSA exercising its power to make changes to the RPI. So for example, the Jevons formula may be used instead of the Carli formula, not just for clothing but more generally.
Mr. Herberg recognises that in principle the Authority could lawfully devise a new index incorporating solutions for each of these deficiencies identified in the RPI and use the power to make fundamental changes to the RPI to adopt that index as the RPI. In my judgment, where the current RPI suffers from extensive flaws, the UKSA is entitled to remedy the problem in that way rather than make piecemeal changes. The same applies to making a change in order to improve the quality of an existing RPI index. For example, the Authority might judge that the use of the new index would be statistically preferable or more robust and more efficient. Of course, the UKSA would need to be satisfied that, in their judgment, the RPI as altered would provide a proper measure of consumer price inflation complying with s.7 of the SRSA 2007, and which they would be able officially to endorse, maintain and publish under s.21(1) and (4).
There is nothing in the language of the SRSA 2007 to prevent the UKSA from changing the RPI by adopting a feature from another index already in existence, whether in the UK or abroad. Moreover, there is no rational reason why Parliament should be content to allow the UKSA to make fundamental changes to the RPI, but not to use solutions to be found in other existing indexes. One major advantage of being able to use an existing index, is that there will often be evidence and analysis as to how that existing index (or components thereof) has performed in practice. There will also be a time series of rates of inflation. It is impossible to see why Parliament should have wanted to deny the UKSA the use of such material.
But Mr. Herberg argues that it is ultra vires for the UKSA to incorporate the methodology and data sources of the whole of an existing index into the RPI. If it is permissible to change the RPI by devising a new index, in order to overcome statistical flaws, I do not see why that power to change does not also apply to the use of an index which already exists. In the first case the RPI has not ceased to exist; instead, it has been “fundamentally changed” to the newly devised index. The same is no less true in the second case.
Not surprisingly, the SRSA 2007 does not indicate that the UKSA may only make one fundamental change at a time. Such changes may be interrelated and it would make no sense to insist that they be made separately. In my judgment the UKSA is entitled to make all the fundamental changes it considers appropriate together, even if that means relying upon a different index.
I also accept the submission of Sir James Eadie that the legislation does not require the UKSA to make an individual assessment of each element of an index it proposes to adopt in order to be satisfied that each is better (or no worse) than the one it replaces. The UKSA is entitled to exercise its judgment by making an overall assessment of whether a new index should be adopted. This issue only arose because the claimants pinpointed a small technical issue in the evidence of Mr. Athow to do with the mathematics of “chain-linking” (WS paras. 112-113). From a “theoretical perspective” it would be preferable to link the annual update of the basket to the ongoing index series once a year (as in the case of the RPI), rather than twice a year (as in the case of the CPIH). But the CPIH eliminates bias, the impact is negligible, and there was no user concern on this issue (see the consultation exercise). Given the claimants’ acceptance that the UKSA can make fundamental changes to the RPI, Sir James Eadie understandably referred to their reliance upon the chain-linking point as “straining out a gnat”.
For all these reasons, the decision to import the methods and data sources of the CPIH was not prohibited by, or ultra vires, s.21 of the SRSA 2007. Ground 1 must be rejected.
Ground 2
Ground 2(1)
The claimants submit that the UKSA acted unlawfully by directing itself that it could only take into account “statistical matters” and so did not take into account the impact of the RPI decision on legacy users, that is the holders of RPI-linked gilts and bonds and persons entitled to RPI-linked pensions. The claimants describe that factor as an “obviously material consideration” which the UKSA was obliged to take into account, applying the principles in R (Hurst) v HM Coroner for Northern District, London [2007] 2 AC 189 at [57].
The claimants submit that the objective in s.7(1) of the SRSA 2007 requiring an “official statistic” to “serve the public good” supports ground 2(1) and, in particular, their contention that the ambit of the UKSA’s statutory functions is not restricted by s.7(3) to “quality” and other purely statistical matters.
The claimants also submit that the veto mechanism in s.21(2) to (4) does not indicate that wider policy issues and the interests of particular groups or individuals are not a matter for the UKSA but for others. The veto provision has been inserted for the narrow purpose of dealing with the potential effects of the early redemption clauses in old-style gilts being exercised. Accordingly, those provisions have no implications for the scope of the UKSA’s functions generally or its functions in relation to the RPI.
Mr. Herberg submits that under the regime which preceded the SRSA 2007 the Chancellor was able to take into account the impact on different parts of society when asked to approve a change to the RPI. Under the SRSA 2007 the Chancellor does not take part in decisions on how the RPI should be compiled. That is a matter for the UKSA. His power is limited to vetoing changes of a certain kind proposed by the UKSA and, in practice, even that power expires in 2030. Accordingly, Mr. Herberg submits that it would be most surprising if in transferring responsibility for maintaining the RPI to the UKSA, neither the Authority nor anyone else has responsibility under the SRSA 2007 for considering the wider effects of changes to the RPI. The same issue arises in relation to changes to other statistics prepared by the UKSA.
Mr. Herberg also submits that the UKSA prevented itself from considering the interests of “legacy users”, not only on the issue of whether the CPIH should be incorporated into the RPI as a matter of principle, but also on how the RPI decision should be implemented, including the timing of any change after 2030. The power to change the RPI includes a power to consider how that change should be implemented.
Discussion
From the earlier review of the evidence before the court it can be seen that, in summary, the UKSA has taken the following approach: -
From about 2011 onwards the UKSA has been concerned about flaws in the statistical reliability of the RPI;
In 2013 the Authority decided to maintain the RPI as a legacy index in order to maintain continuity for certain users, that is those holding index-linked gilts, bonds and pensions. But it did not consider how other parties would be affected by continuing to maintain the RPI without correcting its flaws;
In 2013 the UKSA cancelled the RPI’s designation as a National Statistic. The Authority developed the CPIH and in 2017 designated it as a National Statistic;
From 2013 onwards the UKSA had urged users to move away from relying on the RPI, but 6 years later its use still remained widespread;
By 2019 the UKSA accepted that action needed to be taken to address all the flaws in the RPI which had been identified by then;
The UKSA was plainly aware that alignment of the RPI with the CPIH would have major consequences for a broad range of people and institutions. That reflected the “winners and losers” points made in the report of the House of Lords Economic Affairs Committee. But the UKSA took the view that those wider issues could not affect the discharge of its functions under the SRSA 2007, in particular whether the RPI should be changed and if so how.
Mr. Herberg accepts that the implications of ground 2(1) go way beyond the interests of the holders of RPI-linked investments in this case. Firstly, he agrees that if the effect of the RPI decision on those interests is an “obviously material consideration” which the UKSA is required to take into account, the same would also apply to the effects of both that decision and maintaining the status quo for other institutions, groups and individuals affected, whether winners or losers. Secondly, that conclusion would apply not just to the UKSA’s functions under s.21, but also to the discharge of its responsibilities, including the application of s.7 of the SRSA 2007, in relation to other “official statistics”, for example statistics produced by the Authority. Plainly, therefore, it is necessary to look at the issue more broadly.
Essentially the underlying problem concerns the use to which an official statistic is put. For example, in relation to the RPI the UKSA has no control over the circumstances in which the RPI may be used, whether for the terms of a gilt, student loans, commuter fares, regulation of utility returns or prices, or adjustments to tax regimes. The use of the RPI in each of these areas is likely to give rise to winners and losers. The choice of an index, and the implications of that decision for those affected, are policy issues for the person or body making the decision. The SRSA 2007 gives the UKSA no role to play in determining such policy issues.
Where a public authority or decision-maker has chosen to use the RPI in a particular context the claimants’ submission could be repeated that that party has no influence over the compilation of the index. But it does not follow that Parliament has made the UKSA responsible for taking into account and evaluating all policy issues or consequences for that use of the RPI which would result from its proposal to change the index.
The FDA case illustrates the point. There, a judicial review was brought against the decisions of the Secretary of State for Work and Pensions and HM Treasury to change the basis for the annual indexation of certain social security benefits, state pensions and public service pensions from the RPI to the CPI. If, however, the departments had carried on using the RPI as the basis for that annual indexation, the consequence of the claimants’ legal argument would be that before the UKSA could make a change to the RPI, for example, by substituting the Jevons formula for Carli, resulting in a reduction of the annual indexation by 0.7 of a percentage point, it would be obliged to take into account the effects of that change on the groups of the population or individuals who would be affected negatively by it as the recipients of benefits or pensions. If the claimants are correct, it is difficult to see how the UKSA’s legal obligations could stop there. For example, savings in public expenditure would be equally relevant.
The claimants’ argument would also apply to official statistics produced by the UKSA (s.6(1)(a) and s.20) and the Board’s functions under s.9. Because the claimants base their argument upon s.7, in particular “serving the public good”, its reach would not stop there. Potentially it would apply to the UKSA’s oversight of official statistics produced by, for example, a government department with policy responsibilities for the subject-matter to which its statistics are relevant.
In my judgment, there is nothing in the legislation to indicate that Parliament intended to impose such far-reaching responsibilities on the UKSA. Instead, the Authority has been established as an independent, expert body with specific responsibilities for official statistics. It is directly accountable to Parliament for the manner in which it discharges its functions under the SRSA 2007. The Authority is not accountable to a Minister with policy responsibilities. The Supreme Court has emphasised the fundamental importance in our constitution of the accountability of Ministers to Parliament for policy issues (R (Miller) v The Prime Minister [2020] AC 373 at [46]).
The claimants’ argument would require the UKSA to identify the different winners and losers as the result of a proposed change to the RPI (or other statistics), to evaluate the likely effects upon them and then to weigh the competing interests, before deciding whether to go ahead with the proposal at all, or in an amended form. The same would apply to Mr. Herberg’s suggestion that in the present case the UKSA ought to have considered whether to postpone the change to the RPI beyond 2030 in order to mitigate further its negative impacts upon holders of RPI-linked investments. That delay, while beneficial to institutions and individuals in that category, would be harmful to others who would continue to pay higher prices or interest charges linked to an unaltered RPI.
Such policy conflicts would be unavoidable and yet Parliament has not established the UKSA as a body appropriate or equipped to be responsible for such matters.
I accept Sir James Eadie’s analysis of s.7. The focus of this provision is on the quality or qualities of official statistics, not some freestanding responsibility for the public interest. So the Authority’s objective under s.7(1) is to promote and safeguard the production and publication of “statistics that serve the public good”. Section 7(2) defines “serving the public good” as including “informing the public about social and economic matters” and “assisting in the development and evaluation of public policy”. Section 7(2) expressly states that these expressions only apply in the context of s.7(1). In other words, they simply form part of the UKSA’s statutory objective in relation to statistics. Section 7(2) recognises how statistics may serve as a tool in that respect. But the legislation expressly does not make “the development and evaluation of public policy” one of the objectives of the UKSA, let alone the determination of public policy or the resolution of wider policy issues.
In accordance with the objectives in s.7(1) and (2), s.7(3) imposes a duty on the UKSA to promote and safeguard the quality and comprehensiveness of official statistics and good practice. Section 7(4) elaborates on what is meant by “quality”. Again these provisions confer no function on the Authority of evaluating consequences involving broader policy decisions.
Section 7 of the SRSA 2007 is not concerned with weighing the cost to companies or private individuals, or for that matter benefits to the public purse, resulting from changes to an “official statistic” such as the RPI. The introduction of wider policy issues or the interests of particular groups into the decision-making of the UKSA would distort, if not undermine, the independence and objectivity with which it is required to discharge its functions under the legislation.
The Codes of Practice adopted under s.10 of the SRSA 2007 are entirely consistent with this analysis. For example, principle 1 of the 2009 Code required official statistics to meet “the requirements of informed decision-making by government, public services, business, researchers and the public”. In other words, such statistics may assist others as a source of information when they make their decisions. Meeting such requirements does not imply that the UKSA has any function regarding the making of those decisions or policy choices. Put simply, principle 1 recognises that UKSA’s objective is not to produce statistics as an end in itself, or purely for statistical or mathematical purity. An Official Statistic is used in a practical context and is expected to be fit for its relevant purpose, for example, quantifying a particular factor or expressing a trend.
Here the UKSA was essentially concerned with whether the RPI is a good, or even adequate, measure of consumer price inflation, or the cost of living. Whether going beyond that it is also a good index to be used in a commercial contract such as a lease, or in an investment such as a gilt, are matters for the parties involved in those transactions. Hedging or matching indexed liabilities is likewise a matter for the financial institutions and any regulator involved. There is no legal requirement for the UKSA to have regard to such matters. Likewise the UKSA has no responsibility for the terms upon which gilts are offered. Any question as to whether the Government should continue to borrow money using gilts linked to the RPI, or some other measure of inflation, is a matter for the Chancellor and the Debt Management Office, not the UKSA. These proceedings are not concerned with any decision on that subject by HM Treasury or that Office.
Accordingly, I do not consider that the UKSA made any error of law in stating that it could only take “statistical matters” into account when it reached a decision on whether to change the RPI. Read properly in context I understand “statistical matters” to include not only purely mathematical issues but the fitness of the RPI to be relied upon by users up and down the country as a measure of consumer price inflation, or the cost of living.
The claimants have never grappled with the legal and practical implications of their arguments under ground 2(1). In particular, they have not explained how the language of the SRSA 2007 reveals that Parliament intended the UKSA to have a power, let alone a duty, to resolve competing interests such as we see in the case of the RPI, or indeed more widely. On the correct interpretation of the SRSA 2007, the UKSA rightly decided that the effects of their RPI decision on pension schemes and pensioners, or indeed on other groups in the country, whether winners or losers, fell outside their statutory functions. Those matters were not relevant considerations, let alone “obviously material considerations”.
As for the claimants’ submission that on this interpretation of the SRSA 2007 no one has responsibility for considering the wider effects of changes to the RPI (or another UKSA statistic), there are at least two answers. Firstly, decisions on whether to employ a particular statistic, such as the RPI, in commercial transactions or in a socio-economic scheme are the responsibility of those who make such arrangements. Secondly, and ultimately, this is a matter for Parliament.
The claimants’ argument in relation to the PSED concerned the differing effects of the change to the RPI in relation to women. Mr. Herberg rightly accepted that this contention falls away if, as I have decided, the main argument under ground 2(1) fails (see also R (Adiatu) v HM Treasury [2021] 2 All ER 484 at [209] and [239]-[244]).
For these reasons, ground 2(1) must be rejected.
Ground 2(2)
During the consultation carried out in 2020 a number of holders of index-linked gilts suggested that compensation should be paid by the Government to mitigate losses sustained as the result of the decision to bring the methods and data sources of the CPIH into the RPI. This was addressed at paras. 81-83 of the Response to Consultation published on 25 November 2020: -
It is clear to the government that the Authority’s approach to reform the RPI will have widespread impacts, with the scale and direction of these impacts differing by user. As noted, the holders of index-linked gilts maturing after the date of implementation face reduced returns and the vast majority of index-linked gilt holder respondents called for reform to be implemented as late as possible. In addition, virtually all of these same respondents called for further mitigation in the form of compensation.
As outlined in his 23 October 2020 letter to the Authority Chair, having considered the responses to the consultation (in light of the relevant impacts he could consider) in order to minimise the impact of reform on index-linked gilt holders, the Chancellor has noted that he will be unable to offer his consent to the Authority’s proposal before the maturity of the final specific index-linked gilt in 2030.
On the calls for further mitigation, the government will not offer compensation to the holders of index-linked gilts. The contractual terms of all index-linked gilts state that the RPI should be used to determine the index ratio, which is used to calculate interest and redemption payments. There is no change to this flowing from the implementation of the Authority’s reform.”
The claimants submit that the Chancellor’s decision on this subject was flawed because he considered only the contractual terms of gilts issued from 2002 onwards and did not consider the wider effects of the RPI becoming aligned with the CPIH. They also submit that the Chancellor failed to comply with the PSED.
Discussion
It is plain that the Chancellor had well in mind the wider effects of the UKSA’s decision to change the RPI when he exercised his veto under s.21(3) of the SRSA 2007 in order to reduce negative impacts on those holding index-linked investments (see paras. 10-11, 58-80 and 103-4 of the Response to Consultation).
In addition, on 19 August 2020 the Chancellor was given a briefing on the merits of the compensation sought in the representations made by consultees. This was summarised as follows:-
“In consultation responses, investors have argued that compensation should be offered by redefining the terms of all index-linked gilt contracts such that the reference rate used is equal to the reformed RPI plus a compensatory ‘wedge’. This ‘wedge’ would be either partially or wholly equal to the long run difference between an unreformed RPI and CPIH. Respondents to the consultation have suggested that this ‘wedge’ could be determined through another consultation.” (emphasis added)
Thus, the focus of those representations was that the terms of index-linked gilts should be altered so as to preserve a rate of return above the RPI aligned to the CPIH. The “compensatory wedge” referred to that new form of protection.
In summary, the Chancellor was advised that there were no grounds to offer compensation, whether in terms of justice and equity, wider policy considerations or any legal entitlement (see Ms. MacDougal WS at paras. 68 – 74 and the submission by officials). On 29 August 2020 the Chancellor accepted that advice and decided in principle that compensation should not be paid. It is therefore plain that, contrary to the claimants’ submissions, the Chancellor’s decision to refuse compensation was not based simply upon the contractual terms of the gilts. I also note that the briefing in August 2020 also raised the need to consider the application of the PSED.
On 24 September 2020 the Chancellor was provided with further briefing accompanied by a PSED assessment which included the following:-
“We know from the Pensions Policy Institute’s response to the consultation that women will experience the greatest reduction to their lifetime and annual benefits owing to their on average longer life expectancies. Many of this cohort of women are likely to be also be (sic) affected to (sic) changes to the State Pension Age ….. ”
On 10 November 2020 the Chancellor was given further advice about the broader impacts of the RPI reforms from 2030 on affected groups, including defined benefit pension schemes, indirect taxes, regulated rail fares, and the calculation of student loan payments. This included the fact that average loss of pension earnings for women would be expected to be greater than for men due to the former’s longer average life expectancy. Officials advised that, while the Chancellor should acknowledge the impact of the RPI decision on some defined benefit pension scheme members, he should decline to provide any ‘mitigation’ or compensation to them. On 23 November the Chancellor decided to proceed on that basis.
Reading this material together with Ms. MacDougall’s witness statement at paras. 74-86, it is apparent that the Chancellor was provided with careful, ample briefing on the issues raised by ground 2(2) and that he did not fail to comply with the PSED or to have regard to any material consideration. The fact that the differential effect of the RPI decision on female holders of defined benefit pensions was not drawn to the Chancellor’s attention until after his decision in principle that compensation should not be paid (on 29 August 2020) does not vitiate his “compensation decision” announced in the Response document on 25 November 2020. It is clear that the Chancellor continued to take the same view on the payment of compensation after he had regard to that consideration. It was referred to expressly in para. 97 of the Response document. The claimants are incorrect in describing this as an impermissible “rearguard action”.
The weight to be given to the various matters raised on the wider effects of the change in the RPI were for the Chancellor, subject to judicial review applying the Wednesbury principle. There is no basis for the court to say that the judgment reached by the second defendant was irrational or flawed by any error of law. The claimants’ complaint appears to have been triggered by what they perceive as the inadequacy of the explanation given in the Response to Consultation document on the compensation issue. But the material which has been shown to the court demonstrates that the matter was considered carefully and lawfully. The extent of the reasoning given in the Response to Consultation does not provide any proper basis for the court to intervene.
For these reasons ground 2(2) must be rejected.
Ground 3
Ground 3(1)
The claimants submit that in March 2020 the UKSA failed to consult the public on its RPI decision. The Authority merely consulted on how the methods and data sources of the CPIH would be brought into the RPI, while the Chancellor consulted on whether to consent to the change to the RPI being brought into effect between 2025 and 2030. The claimants say that the UKSA was under a legal obligation to consult on the RPI decision on three alternative bases:-
The UKSA did in fact embark upon a process of consultation and it was therefore obliged to do so properly (R v North and East Devon Health Authority ex parte Coughlan [2001] QB 213 at [108]) which required compliance with the four “Gunning Principles”;
A duty of proper consultation also arose as a matter of fairness from “the pressing and focused impact” on the claimants and other stakeholders (R (Bhatt Murphy) v Independent Assessor [2008] EWCA Civ 755 at [49]);
Consultation was required for the UKSA to satisfy its duty of sufficient inquiry (Secretary of State for Education and Science v Tameside Metropolitan Borough Council [1977] AC 1014, 1065A-B).
If the UKSA was obliged to consult on the RPI decision, the claimants complain that:-
No consultation took place on the issue of whether the RPI should be maintained when that proposal was at a formative stage, or indeed at any other time. When the joint consultation was launched in March 2020, the UKSA had already made that decision;
In the Response document the UKSA did not conscientiously take into account the representations made on behalf of legacy users such as the claimants.
There is no factual dispute about point (i).
Mr. Herberg stated that the claimants do not rely upon any substantive legitimate expectation. They accept that, subject to grounds 1 and 2, the UKSA was entitled to take the RPI decision, provided that proper consultation took place beforehand.
The claimants do not rely upon any promise or established practice of consultation. This is relevant because of the consultations which were carried out between 2012 and 2016. To qualify as a legitimate expectation, a practice must be clear, unequivocal and unconditional; it must be sufficiently settled and uniform to give rise to an expectation that the claimant would be consulted (Bhatt Murphy at [29] and R (Brooke Energy Limited) v Secretary of State for Business, Energy and Industrial Strategy [2018] EWHC 2012 (Admin) at [53]). Instead, the claimants are relying upon “the secondary case of procedural expectation” (Bhatt Murphy at [49]).
Discussion
It is necessary to be clear about the ambit of the failure to consult to which the claimants’ complaint relates. On the Tameside argument, the claimants say that the UKSA needed to know the consequences of the RPI decision for users of that index, in particular legacy users. They rely upon a statement by the Chancellor that the Treasury did not have sufficient information about the effects on such users. In relation to consultation based upon a duty to act fairly, it is relevant for the court to consider the nature of the representations which the claimants say they were unable to have taken into account by the UKSA (R (Plant) v Lambeth London Borough Council [2017] PTSR 453 at [86]-[88]). In this regard, and consistently with the way in which the claimants seek to rely upon Bhatt Murphy, Mr. Herberg said that the consultation by the UKSA should have covered, and did not cover, the impact of the RPI decision on the claimants, their stakeholders and others in a like position.
Thus, the scope of the deficiencies in the consultation upon which the claimants rely begs the question whether those matters were relevant to the statutory functions of the UKSA. Consequently, Sir James Eadie submitted that the consultation challenge must fail if it relates to matters falling outside the scope of those functions. I understood Mr. Herberg to accept that proposition. I certainly agree with it. In view of the reasons I have given for rejecting ground 2(1), it must follow that ground 3(1) fails on this basis alone. However, I will deal briefly with other issues which have been argued.
In relation to ex parte Coughlan, it is plain that the UKSA did embark upon consultation on how the methods and data sources of the CPIH would be brought into the RPI. But I do not understand that the Gunning principles on carrying out a consultation properly may be used to contend that the scope of that consultation should have been wider. The Gunning principles apply within the ambit of the consultation being undertaken. Mr. Herberg did not cite any authority to indicate otherwise.
In relation to the secondary case of procedural expectation, the claimants placed a good deal of emphasis upon the following passage in the judgment of Laws LJ in Bhatt Murphy at [49]:-
“….. Accordingly for this secondary case of procedural expectation to run, the impact of the authority's past conduct on potentially affected persons must, again, be pressing and focussed. One would expect at least to find an individual or group who in reason have substantial grounds to expect that the substance of the relevant policy will continue to enure for their particular benefit: not necessarily for ever, but at least for a reasonable period, to provide a cushion against the change. In such a case the change cannot lawfully be made, certainly not made abruptly, unless the authority notify and consult.”
But it is essential to consider the preceding analysis from which that passage was derived. At [40] Law LJ posed the question:-
“… what are the conditions under which a public decision-maker will be required, before effecting a change of policy, to afford potentially affected persons an opportunity to comment on the proposed change and the reasons for it where there has been no previous promise or practice of notice or consultation?”
At [41] Laws LJ set out overarching principles. The secondary case of procedural expectation is concerned with an exceptional situation, “a long way distant from the archetype of public decision-making”. The law will not often require a public authority to involve a section of the public in its decision-making process by notice or consultation if there has been no promise or practice to that effect. There is an underlying reason for this. Public authorities typically, and central government par excellence, enjoy wide discretions which it is their duty to exercise in the public interest. They have to decide the content and pace of change. Often they must balance different, indeed opposing, interests across a wide spectrum. Generally they must be the masters of procedure as well as substance. This entitlement to formulate and re-formulate policy is ordinarily repugnant to an enforced obligation, in the name of a procedural legitimate expectation, to take into account and respond to the views of particular persons whom the decision-maker has not chosen to consult. At [42] Law LJ stated that for a secondary case of procedural expectation, “something no less concrete [than an express promise or an established practice] must be found.”
I agree with Sir James Eadie that the secondary case of procedural expectation does not apply to the circumstances relied upon by the claimants. The expression “pressing and focused” appears in the judgment of Laws LJ at [49] and [58], but in both places that term refers back to what he had said at [46]. Consequently, although there may theoretically be no limit to the number of potential beneficiaries of a legitimate expectation, in reality it is likely to be small for two reasons. First, it is unlikely that such an expectation will be recognised if it is said to apply generally or to a diverse class. Second, the broader the class the more likely it is that a supervening public interest may justify a change of position. In the secondary case where there has been no previous promise or practice of consultation, those limitations apply just as forcefully.
In the present case, the effects of the RPI decision are general or diverse. That is because of the wide range of circumstances in which the RPI continues to be used in practice. Although the financial impact of the change in the RPI on financial institutions and pensioners is very substantial, it applies to a large number of commercial bodies and individuals affected by the use of RPI for indexation. The evidence before the court shows that there are many other companies and individuals who may benefit from the upwards bias of the RPI. But any obligation of a public authority to consult would be owed not only to those expected to oppose its proposal, but also those who might support it because, for example, they would benefit from it. Thus, it is plain that the overall impact of a proposal to align the RPI with the CPIH would be very diverse, not focused. The fact that a consulting party would have a duty to balance a number of competing interests across a wide spectrum militates strongly against the application of the Bhatt Murphy exception (see Brooke Energy at [70]).
In addition, because it must have been apparent from 2012 onwards to the claimants and others in the like position that maintaining the RPI without correcting its flaws would be adverse to many “losers”, I do not accept that the claimants had substantial grounds to expect that those flaws in the RPI would continue to enure for their particular benefit. What if the mathematical flaws in the RPI had biased the index so that it was consistently underestimating consumer price inflation? Parties who would have benefited from those uncorrected flaws would not be entitled to be consulted on the effects of losing that benefit, relying upon the Bhatt Murphy exceptional case, before the UKSA could correct those flaws.
The duty of sufficient enquiry is concerned with the rationality of a defendant’s decision, not consultation based upon fairness (see e.g. R (Balajigari) v Secretary of State for the Home Department [2019] 1 WLR 4647 at [70]). The claimants have come nowhere near establishing irrationality. This argument would only arise if the wider effects of the RPI decision were relevant to the UKSA’s functions, contrary to the conclusions I have reached under ground 2(1). But on that basis, those wider effects would have to include not just the claimants and their stakeholders, but potentially all parties affected by the proposed change to the RPI, whether positively or negatively. This takes one back to the impossible task with which the UKSA would then be faced of weighing competing interests in diverse areas, where they have no policy-making responsibility or capacity. When dealing with ground 2(1) Mr. Herberg accepted that the UKSA would be entitled to say that they could not carry out any such weighing exercise (transcript day 2 p.172). Although, at that stage in the argument he was making the point that the UKSA had not reached that conclusion, his submission nonetheless reflected the insuperable difficulty which, in reality, the claimants face in establishing irrationality.
For all these reasons ground 3(1) must be rejected.
Ground 3(2)
The Chancellor did not consult on the issue of compensation for legacy users. Nonetheless, those users raised the matter in a number of substantial representations. As we have seen, the Chancellor was briefed on that subject and he accepted the advice that compensation should not be paid. I have already rejected ground 2(2).
The claimants have not demonstrated any legal basis for asserting that the Chancellor was under a duty to consult on whether compensation should be provided to legacy users as a result of the RPI decision.
However, assuming that a duty of consultation did arise, the claimants’ complaint is that the Chancellor did not include sufficient reasoning or information to enable legacy users to give intelligent consideration and response. It is not clear from the claimants’ submissions what specifically they claim to have been required on this aspect, absent which legacy users were deprived of a proper opportunity to make representations. The proof of the pudding is in the eating. I have reviewed the extensive and intelligent representations made by legacy users and their representatives. They were able to say what the effects on them would be and what compensation they considered should be paid. I was not shown anything to explain what additionally legacy users would have wanted to know from the Treasury, or what additional representations they were prevented from making to the Chancellor.
For all these reasons, ground 3(2) must be rejected.
Delay
For the reasons given previously, I have no doubt that the RPI decision, the decision on the principle of bringing CPIH methods and data sources into the RPI, was taken on 14 February 2019. That decision and related documents were well publicised when the announcements on the consultation to be carried out were made on 4 September 2019. Those announcements made it clear that the consultation would be on (i) whether the Chancellor would consent to the change being brought into effect between 2025 and 2030 and (ii) the method by which the UKSA would make the change.
It was plain from the documentation published on 4 September 2019 that the consultation would not include the principle of the UKSA’s decision that the change to the RPI should be made, whether before or after the expiration of the Chancellor’s veto in 2030. It was also obvious that the Chancellor was not proposing to consult on the payment of compensation. Anybody who was concerned about the legality of the ambit of the consultation exercise could and should have sent a pre-action protocol letter at that stage. If they considered the response given to their complaint unsatisfactory, they should then have started proceedings for judicial review. I was given no good reason as to why that straightforward course was not taken.
Turning to the consultation document issued on 11 March 2020. I do not think the analysis is materially different, notwithstanding, for example, paragraphs 7, 40 and 65.
The concern that the claimants and others would have had about the scope of the consultation exercise went to a fundamental matter: they say that the RPI decision had already been taken, with the legal consequence that the first Gunning principle (that the consultation should be on a proposal at a “formative stage”) could not be satisfied. That indeed became the claimants’ case in this court (see e.g. para. 94 of the claimants’ skeleton). Unless that alleged flaw was raised by the claimants or consultees in September 2019 or in March 2020, it was not a matter which could safely be left to the “self-correcting nature” of the consultation process actually being undertaken at the time. Unless it was raised, and indeed “corrected” by UKSA (assuming that the complaint had merit), it would be “irretrievable” by the end of that consultation process in November 2020 (see R (Royal Brompton and Harefield NHS Foundation Trust) v Joint Committee of Primary Care Trusts [2012] EWCA Civ 472 at [87]-[93]). In my judgment, this analysis is not materially affected by passages in the March 2020 Consultation document (see e.g. [112] above) which the claimants suggested “left the door ajar” on the scope of the consultation. If anything, those passages simply reinforce the need for anyone concerned about that issue to have raised it then and not wait until the outcome of the consultation exercise. The position is no different on the compensation issue.
Accordingly, the period of delay before the claim was brought is very considerable. The claimants ask for an extension of time. I do not consider that the court has been given adequate reasons to justify that delay. Moreover, there would be detriment to good administration through the consultation process having to be repeated. The question also arises what substantial benefit could that achieve? It has not been suggested that any different view should be taken by the UKSA on the flaws which exist in the RPI, such that the index is a poor measure of inflation with an ongoing upwards bias. Leaving to one side the conclusion I have reached on ground 2(1) of this challenge, Mr. Herberg fairly and properly accepted that in a re-run of the consultation, the UKSA would be entitled to say that they are in no position to evaluate and weigh the competing interests of the winners and losers of leaving the RPI as it is or aligning it with the CPIH.
The parties emphasised to the court that its decision in this case would be market sensitive. Equally the decisions announced on 4 September 2019 and 11 March 2020 were market sensitive (see e.g. the briefing to the Chancellor dated 6 September 2019 and 10 November 2020). That only serves to reinforce the conclusion I have reached that any challenge to the scope of the consultation should have been brought at an early stage.
When granting permission to apply for judicial review Lang J left the issue of delay to the trial judge. Although I have already rejected the relevant grounds of challenge on the merits in any event, I also refuse the extension of time sought by the claimants in relation to ground 3.
The private law claim
This claim arises if, as I have decided, the RPI decision is held to be lawful (see the title to Section G of the Statement of Facts and Grounds). In those circumstances, the claimants submit that the cessation clause in gilts issued from 2005 onwards will be triggered when the change to the RPI is implemented at some point after February 2030. Their contention is that the effect of the UKSA’s decision is not merely to amend the RPI but to abandon it in favour of the CPIH, thereby resulting in the RPI “ceasing to be published”. At paragraph 107 of the Statement of Facts and Grounds the claimants state that the analysis to support that conclusion is the same as that relied upon in support of ground 1.
Already a difficulty in the claimants’ case becomes apparent. The claimants only need to rely upon the cessation clause if ground 1 fails. But that ground fails if the RPI decision is intra vires, so that implementing that decision will not result in the UKSA breaching its duty under s.21(1) of the SRSA 2007 to publish the RPI. Yet the compensation clause would only be triggered if the RPI ceases to be published. Why should that question, apparently the same as the one raised by s.21(1), be answered any differently, particularly if the claimants are relying upon the same analysis as they advanced under ground 1?
Both parties agree that the issue for the court is one of the proper construction, and then application, of the cessation clause. In this context, principles of contractual interpretation are relevant, rather than statutory construction. The parties referred to Wood v Capita Insurance Services Limited [2017] AC 1173 at [10] and [12]. In addition, the proper interpretation of the clause should accord with its purpose and business common sense.
The claimants submit that the purpose of the clause is to provide protection to investors in index-linked gilts. Some form of investor protection is required against changes to the RPI or its replacement by a different index. The claimants go so far as to contend that the protection provided by the cessation clause relates to “the substantive rate of return as linked to the measure of inflation”. On that basis they say that the relevant question under the cessation clause is whether that measure of inflation continues or whether it will cease to be published (claimants’ skeleton para. 112). Thus, the claimants’ case is concerned with the protection of a rate of return. Their submission to this court echoes the representations to the Chancellor on the “compensatory wedge” (see [183] above).
Nevertheless, Mr. Herberg accepted that a “fundamental change” to the RPI is insufficient to trigger the cessation clause. In that event, the RPI does not cease to be published, whether for the purposes of s.21(1) of the SRSA 2007 or the cessation clause.
In summary, Sir James Eadie submitted that if the RPI is treated as continuing to be published for the purposes of s.21 of the SRSA 2007, then it is also to be treated as continuing to be published for the purposes of the cessation clause, properly construed. The cessation clause only addresses the situation where the RPI ceases to be published. In a case where the clause is triggered, an alternative index has to be designated. The clause does not protect investors against changes in the RPI generally, nor does it seek to protect any particular rate of return or provide for the payment of any compensation to stockholders.
Both sides seek declaratory relief giving effect to their respective positions.
Discussion
The parties asked the court to treat the protective clauses which preceded the cessation clause as part of the factual matrix for interpreting the latter.
The regime governing alterations to the RPI is also a relevant part of the matrix. The UKSA was not established until the SRSA 2007 came into force. So gilts issued before then with the 2005 cessation clause, as well as old-style gilts and gilts with the “2002 clause” (see below), were purchased on the basis of the RPI regime which existed before the 2007 Act. At that stage the coverage and methodology of the RPI was a matter for the Chancellor to determine exercising his common law powers as a Minister of the Crown. He was not subject to any statutory regime. There was no statutory duty to maintain and publish the RPI.
The protective clause in the old-style gilt read as follows: -
“If any change should be made to the coverage or the basic calculation of the Index which, in the opinion of the Bank of England, constitutes a fundamental change in the Index which would be materially detrimental to the interests of stockholders, Her Majesty’s Treasury will publish a notice in the London, Edinburgh and Belfast Gazettes immediately following the announcement by the relevant Government Department of the change, informing stockholders and offering them the right to require Her Majesty’s Treasury to redeem their stock. For the purpose of this paragraph, repayment to stockholders who exercise this right will be effected, on a date to be chosen by Her Majesty’s Treasury, not later than seven months from the last month of publication of the old Index. The amount of principal due on repayment and of any interest which has accrued will be calculated on the basis of the Index ratio applicable to the month in which repayment takes place. A notice setting out the administrative arrangements will be sent to stockholders at their registered address by the Registrar at the appropriate time.”
The protection provided in the old-style gilts had the following key characteristics: -
It only arose if the coverage or basic calculation of the RPI was changed so as to represent in the opinion of the Bank of England a “fundamental change in the Index which would be materially detrimental to the interests of stakeholders”;
The clause was not triggered by a change which was not fundamental or materially detrimental to those interests;
When the clause was triggered the investor’s only protection was early redemption. Whether that option was taken up was a matter of choice for each individual. Some investors might choose to continue holding the gilt. As the evidence shows, in some markets that might appear to be a preferable investment decision;
Old-style gilts did not provide for an alternative index to be substituted for the altered RPI;
The protection did not provide for any compensation in the event of the investor choosing not to redeem;
The gilt did not protect any particular level of return, including a return extending into the future.
In 2001 the Debt Management Office carried out a consultation on the terms of new issues of gilts, which included the question whether gilts should be indexed to something other than the RPI. The consultation referred to the research being undertaken at that stage on the methodology of the RPI. In order to bring UK practice into line with that followed in other major markets, the Office proposed to substitute a “replacement index clause” for the early redemption clause.
Consequently, from 2002 new gilts were issued without an early redemption clause, but with the following provision: -
“The value of the principal on redemption will be related, subject to the terms of this Memorandum, to the movement during the life of the Stock of the United Kingdom General Index of Retail Prices (RPI), or any subsequent index that, in the opinion of the Chancellor of the Exchequer after consultation with a body that the Chancellor of the Exchequer considers to be independent and to have recognised expertise in the construction of price indices, continues the function of measuring changes in the level of UK retail prices. The selection of the new index by the Chancellor of the Exchequer shall be conclusive and binding on all stockholders.”
This clause provided for the Chancellor to select a “subsequent index” that would continue “the function of measuring changes in the level of UK retail prices”, after having consulted with an independent, recognised expert body. The clause did not contain any trigger based upon a change to the RPI, fundamental or otherwise. It did not provide for any compensation to be paid to stockholders. The only protection was the selection of a “subsequent index” with a function described in very broad terms. On that basis the “subsequent index” might be somewhat different from the RPI. The clause did not protect any particular rate of return.
Both Mr. Herberg and Ms. MacDougall (WS at para. 22) say that the 2002 gilt concerned some investors because it was thought that a “subsequent index” might be designated by the Chancellor as the index to which the gilt would thereafter be linked even though the RPI continued to be published.
In order to address that concern, gilts from 2005 onwards provided protection through a “cessation clause”. The gilt would only be linked to a different index if the RPI ceased to be published. In other words, the Chancellor could not switch the inflation linking of a gilt to a different index while the RPI still existed.
From 2005 the cessation clause provided (Footnote: 1):-
“The value of the principal on redemption will be related, subject to the terms of this Offering Circular, to the movement during the life of the Gilt of the Index. The “Index” means the United Kingdom General Index of Retail Prices (RPI), or if the existing Index ceases to be published any replacement Index existing at that date which, in the opinion of the Chancellor of the Exchequer after consultation with a body that the Chancellor of the Exchequer considers to be independent and to have recognised expertise in the construction of price indices (the “Consultant Body”), continues the function of being an officially recognised index measuring changes in the level of UK retail prices. The selection of the replacement Index by the Chancellor of the Exchequer (together with such consequential changes to the calculation of the Index Ratio (as defined below) as the Chancellor of the Exchequer may, after consultation with the Consultant Body, consider to be just and equitable having regard to the interests of Her Majesty’s Treasury (as issuer of the Gilt) and of holders of the Gilt) shall be conclusive and binding on all holders of the Gilt. If a replacement Index is selected in accordance with this paragraph, the Index figure used for the purposes of paragraph 16 will be determined by reference to the replacement Index with effect from the first month for which the existing Index figure is no longer published”
The cessation clause has the following characteristics: -
Essentially it forms part of a provision which defines the index used in determining the value of the gilt on redemption and the “index ratio” for calculating coupon payments;
Like the 2002 clause, the cessation clause is not triggered by a change to the RPI. But the cessation clause expressly provides that it is only triggered where the RPI (or another index which has previously replaced the RPI) ceases to be published, so that the Chancellor cannot link a gilt already issued to a new index while the RPI continues to be published;
Where the clause is triggered, the new index must be one which inter alia “continues the function of being an officially recognised index measuring changes in the level of UK retail prices”;
The cessation clause still does not provide for any compensation to stockholders where a replacement index is selected, nor does it protect any substantive level of return.
Unlike the early redemption clause in the old-style gilts, the cessation clause does not ask whether the RPI has been changed in any respect. Indeed, the claimants accept that the clause is not triggered where a fundamental change is made to the RPI. So if, for example, the RPI were to be amended by substituting the Jevons formula for the Carli formula, the cessation clause would not be triggered. The clause does not provide protection for any given level of return.
A key requirement of the cessation clause is that there should always be “an officially recognised index measuring changes in the level of UK retail prices”. I agree with Sir James Eadie that the purpose of the clause is to enable a replacement index to be selected to fill a gap if the RPI in that sense ceases to exist. The clause was necessary to give confidence to investors that they would not cease to receive any inflation protection for their coupon payments and redemption value because the RPI (or a replacement index previously selected under the clause) had ceased to be maintained and published. That could happen, for example, if s.21 were to be repealed altogether, or amended so that the duty to publish was replaced by a power to publish the RPI (and publication subsequently ceased). Given the long duration of gilts, this protection against inflation provided by the cessation clause accords with business common sense.
On the language, factual matrix and purposes of the cessation clause, I consider that the answer to the question whether the RPI continues to be or has ceased to be published is no different, whether it is given under s.21(1) of the SRSA 2007 or under the cessation clause.
For the above reasons, the Chancellor is entitled to the declaration he seeks.
Conclusion
For the above reasons, (1) the application for judicial review is dismissed and (2) the second defendant is entitled to a declaration that, on the materials before the court in these proceedings, the cessation clause will not be triggered by the implementation of the RPI decision in or after 2030.