On 2 November 2016, the Court of Appeal handed down its decision in Barnardo's & Ors v Buckinghamshire & Ors.
Was the Retail Prices Index (RPI) or the Consumer Prices Index (CPI) the most appropriate index for revaluing deferred pensions and increasing pensions in payment? More importantly, who had the power to decide? In this update, we summarise the key points of the judgment that could have wider relevance to pension scheme trustees and sponsors.
A scheme-specific question
The question of which index to use, and/or who has the power to decide, are scheme-specific questions, dependent on the rules of the scheme.
A body of authorities but no definitive answers
This case will join the cases of QinetiQ and Arcadia as part of a body of authorities on interpreting scheme rules on questions of inflation indices. However, every scheme is different. Because of the potential savings to be obtained from switching away from RPI, disputes over RPI and CPI in pension schemes are likely to continue for some time.
A right to A or B
The Court of Appeal provided clarity on an important point in relation to subsisting rights under section 67 of the Pensions Act 1995. Where the scheme rules contain a power to adopt a different index, exercising that power in respect of pensions already accrued will not infringe the subsisting rights protections of section 67.
Context is important
The specific words in question are not the only things examined in such cases. The correct construction of scheme rules is also influenced by the remainder of the document. It can also be influenced (up to a point) by the background facts.
RPI/CPI - the debate continues
As with many issues concerning UK occupational pensions, the question of how to determine inflation-linked pension increases ought to be simple, but is far from it. The Court of Appeal decision in Barnardo's provides a further court judgment to aid us in questions of whether RPI or CPI is the appropriate index and (more to the point) who has the power to determine it.
Rules of the Barnardo's pension scheme
The judgment confirmed that the question of which index to use, and who has the power to decide it, is a scheme-specific question which is dependent on the correct construction of the rules of the scheme in question. Often, the possibility of this debate occurring will have been far from the draftsman's mind when drafting scheme rules, so, unfortunately, drafting style or accidental choices of language can have very significant unintended consequences.
I will not focus in detail on the rules of the Barnardo's scheme, because the details are of interest only to that scheme and its stakeholders, but the key provision was a definition of RPI as follows:
"Retail Prices Index means the General Index of Retail Prices published by the Department of Employment or any replacement adopted by the Trustees without prejudicing Approval…"
The question the court had to decide was:
- whether the reference to a "replacement" index gave the trustees the power to adopt a replacement index to RPI of their own choosing, even if RPI continued to exist; or
- whether the word "replacement" limited the trustees' discretion to circumstances where RPI itself had been replaced by its publisher.
The court held by a two-to-one majority that the second was the correct interpretation.
Wording like that quoted above is fairly common. However, the judges did not consider it in isolation, but in the wider context of the rules - their preference was influenced by the further use of the word "replaced" later in the relevant provision, which indicated, in that context, that the draftsman was using it to refer to replacement by RPI's publisher. It was also influenced by the drafting used in theappendix to the rules which incorporated the old Inland Revenue limits. The content of those revenue limits themselves was also a factor in construing the language.
The fact that one of the three judges produced a clear and well-argued dissenting judgment shows how there is often not a clear-cut answer to this, and individual lawyers (and judges) can interpret the same words in dramatically different ways.
Although each case turns on the rules in question, they all give us - as lawyers - further material to draw upon when we seek to advise trustees and employers on how their rules might be interpreted if disputed in court. The fact that this is a Court of Appeal judgment gives it added authority. It sits alongside the earlier cases of QinetiQ and Arcadia as part of a body of authority; the more that body grows, the easier it will become to discern general legal principles that can be applied to most forms of drafting.
As things stand, we are still in the early stages: disputes over RPI and CPI in pension schemes are likely to continue for some time.
One point from the case which does have wider application is the Court of Appeal's conclusion on the question of accrued rights.
Rights in a pension scheme that have already accrued (so-called "subsisting rights") are protected from detrimental modifications by section 67 of the Pensions Act 1995. The question therefore arises of what the subsisting right is, in relation to a pension with an associated increase in line with "RPI or a replacement index adopted by the Trustees", or similar wording.
On one interpretation of such language, RPI is the default basis for increases, and if the trustees use their power to adopt an alternative index to increase pension deriving from past service, that is a modification to subsisting rights and therefore subject to section 67.
On the other interpretation, however, the pension accrued on the basis that it would be increased in line with either RPI or whatever index the trustees were using at the time when the increase fell to be applied. Following that logic, the trustees are not affecting any subsisting rights if they do indeed use an alternative index to RPI.
The High Court cases of QinetiQ and Arcadia both chose the latter interpretation. Their reasoning was open to question, not least because it appears inconsistent with the Court of Appeal's decision in Aon v KPMG, which gave rise to similar issues albeit not in the context of pension increases (although the judges in QinetiQ and Arcardia did consider and dismiss that argument).
As a result of Barnardo's we now have Court of Appeal authority upholding the reasoning in QinetiQ and Arcadia. It is now difficult to argue that QinetiQ and Arcadia are inconsistent with the Court of Appeal decision in Aon v KPMG, when the Court of Appeal itself does not appear to think so. Now the point is only likely to be decided differently if a case were to reach the Supreme Court.
One of the judges put the point succinctly: "if a person has a right to 'A or B', one cannot say that he has an accrued right to A. He has an accrued right to one or other of them".
It still remains possible to argue to the contrary. For example, consider the situation of a pension scheme that has a power of amendment that is broad enough to allow retrospective amendments, as many schemes do. That could be characterised as an "A or B" scenario: members have the right to, as a default, the benefits set out in the Rules, or as an alternative, to any lower benefit that may be imposed by retrospective exercise of the amendment power. It is hardly likely that any court would hold that section 67 does not protect the "default" rights in that scenario, and so the "A or B" analysis is perhaps too simplistic to work in all situations.
Be that as it may, it now seems unlikely that there will be any change for the foreseeable future. That does therefore remove one potential hurdle from employers and trustees that are looking to move away from RPI-based increases on past service.
However, as with everything in this area, it depends on the correct construction of the actual words used in that particular scheme.
Employers or trustees that are considering changing the index used for pension increases will need careful legal advice. The consistent theme that emerges from the case law is that the question of what is possible depends upon the correct construction of the rules of the individual scheme, and this is not always obvious. Furthermore, it is not simply a matter of what the words on the page say, as the correct construction is also influenced by the remainder of the document as well as (up to a point) the background facts.
Even in the happy situation where the current rules have clear and unambiguous wording, in reality, the position may not be clear-cut. It is not only the current rules that are relevant: for most schemes, the historical rules also need to be carefully reviewed.
It is therefore as well to get considered legal advice in advance, rather than to go ahead and be unprepared for the complaints that may follow from aggrieved members.
If after having carried out the above checks, it is established that a power to switch from RPI to CPI does exist, there remains the question of whose power it is to make that decision. If it is the trustees' power, would it be a proper use of the power to exercise it in this manner? Determining this requires a consideration of the scheme's circumstances and purposes, to be able to judge whether a switch would further the purposes of the scheme or be contrary to them.
The difference in value between CPI and RPI increases can be immense, so there is a strong incentive from those looking to eliminate deficits to adopt CPI in place of RPI. However, the corollary of that is that such a switch is likely to be resisted by members. It is worth noting that CPI and RPI are not the only available indexes, and it remains to be seen whether there is a trend towards use of an index that sits somewhere in between them.
Unless parliament legislates for a statutory override to enable all schemes to switch to CPI (which many are calling for, but which would be politically quite a brave move), disputes over RPI/CPI are likely to continue. Cases like Barnardo's at least help us to gain a better idea of how courts are likely to resolve such disputes.