"Ignoring Irving Berlin's warning that there may be trouble ahead, Holdings decided to dance", so remarked Warren J in paragraph 56 (out of 1596) in his judgment in IBM.
These were not just random musings by the judge, but a reference to IBM's decision to plough on with its proposal to close its UK pension plans to future accrual of defined benefit pension – which proposal went by the name "Project Waltz" – despite the trustee's request for a postponement in October 2010 when new information about relevant pension rights came to light1. That new information was not determinative of the matters at issue before the judge, but IBM's reaction to it was seemingly characteristic of its approach to the closure to accrual more generally, as I will explain.
What key issues were considered and conclusions reached in the case?
At its heart, the case concerned:
- whether IBM had power to close its plans to defined benefit accrual; and
- if so, whether there was a valid exercise of its power.
There was also an important consideration of the way in which IBM had conducted its consultation about Project Waltz.
The first point: did IBM have the necessary power?
The key issues in relation to the first point were (i) whether the power that IBM purportedly used had been properly introduced into the plan rules2 and (ii) whether, in fact, that power enabled IBM to end all defined benefit accrual as it asserted. The short answers to those issues were, respectively: yes, subject to one limitation (to which I shall return) and yes.
The principles evident from Warren J's decision on those issues are more important for other employers and trustees than the detail of them. They are familiar principles, but ones that bear repeating: you have to look carefully at the provisions of your scheme, read in context, to understand what they mean (as the same words used in someone else's scheme might not necessarily have the same meaning in your scheme) and you cannot always take your provisions at face value (if the introduction of them did not fully comply with the scheme's amendment power, they might, in fact, be subject to an implied limitation or worse still, be completely invalid).
The second of those principles had particular effect in this case. The apparently unfettered power forIBM to end defined benefit accrual had to be read subject to a limitation3 that it could not be used to "break the link" between pension already accrued and any increase in the value of it arising from pay rises awarded after the cessation of accrual. This is due to a – reasonably common – restriction on the plan's amendment power protecting accrued rights4.
There was, however, a twist in the tale. As part of Project Waltz, IBM invited members to sign certain "non-pensionability agreements" by which any future rises in pay would not increase the value of members' past service defined benefit pension rights: the alternative being that the members would receive no pay rise. After some refinement, these agreements applied only to pay rises made between 2009 and 2011, and were not affected by any breach of the "Imperial duty" (see below). Further agreements were then entered into in 2011. The judge did not hear sufficient argument to make any ruling on these, but left the door open for the parties to come back and raise further points on this as part of the "remedies hearing" (see below) should they wish.
The second point: did IBM use its power in a valid way?
There was a technical point made as to whether IBM had used its power to close the plans to future defined benefit accrual for an improper purpose, which turned on the way in which the power at issue was worded and was given short shrift by the judge.
The more significant question was whether, in using its power, IBM was in breach of its "duty of good faith".
Warren J (uncontroversially) stated that the principle is "firmly established" that "there is implied in a contract of employment a term that the employers will not, without reasonable and proper cause, conduct themselves in a manner calculated or likely to destroy or seriously damage the relationship of confidence and trust between employer and employee"5 citing Woods v WM Car Services (Peterborough) Ltd  ICR 666 and Malik v Bank of Credit and Commerce International SA AC 20 as authority for this.
Whilst the principle itself maybe uncontroversial, it has not always been applied in an identical way across cases. Further, the application of a parallel, but not necessarily identical, duty in relation to pension schemes6 has long-since been recognised, but its detail not fully "nailed down" either. So, Warren J grasped the nettle and decided to do so7.
The focus of Warren J's analysis was on the Imperial duty – i.e., the "duty of good faith" as it applies to the exercise of employer powers under pension schemes – as that was the main (although not, as I shall explain, exclusive) application of the good faith duty to the issues at hand.
After some debate, the judge decided that the essence of the Imperial duty is that an employer may not act irrationally or perversely when exercising its powers under a pension scheme8, in the sense of being conduct that no reasonable employer would pursue9. However, even where an employer does so, such conduct would need to destroy or seriously undermine the relationship of trust and confidence that exists between the employer and scheme beneficiaries for it to constitute a breach of the Imperial duty.
It was recognised that, in principle, employers are perfectly at liberty to take account of their own financial interests without breaching the Imperial duty10. However, what is particularly noteworthy about this case is the judge's treatment of the "reasonable expectations" that members had about their pension benefits and how that is factored into an assessment of whether the Imperial duty has been breached. Specifically, where there are such reasonable expectations and an employer engages in (what, viewed in isolation would be) reasonable commercial conduct, the employer will still be in breach of the Imperial duty, if that conduct is contrary to the members' reasonable expectations and is conduct that (in light of those reasonable expectations) no reasonable employer would have embarked on11.
The judgment considers, over 300-odd pages, whether members of the IBM plans had any reasonable expectations that defined benefit pension accrual for them would not be terminated and, if they did, whether Project Waltz breached those expectations in a way that seriously undermined the relationship of confidence and trust, and in a way that no reasonable employer would have done. The short answers are, respectively, they did and it did!
Those expectations had developed as a result of two previous exercises - one in 2004/5, the other in 2005/6 - that resulted in material changes to IBM's defined benefit pension provision and communications issued to members at those times. The communications as to the second exercise (at least) made it clear that no guarantees were being given that the plans would remain open to defined benefit pension accrual. Nevertheless, these confirmations were part of wider communications that, the judge held, were intended to provide assurance to members that accrual would not be ended at least for a time – with there being no precise duration of that commitment, but it lasting at least until 6 April 2011 when defined benefit accrual was ended.
The judge also held that IBM's main reason for ending defined benefit accrual was to help it to meet financial targets that had been set with a view to increasing its earnings per share, to help support / enhance its share price – put another way, to benefit its shareholders in light of a commitment that had been given to that end, rather than because (for example) it could genuinely no longer afford to pay for future defined benefit accrual.
There was some debate around whether the closure to defined benefit accrual was necessary to meet the commitment made to shareholders, but even if it was, the commitment was made afterIBM had engendered the reasonable expectations in members about future pension provision and so no reasonable employer would have embarked on Project Waltz in the circumstances.
The third point: issues over IBM's consultation
The judge also held that IBM was in breach of its duty of good faith over how it consulted aboutProject Waltz – in the sense of the term implied into employment contracts, rather than (this time) the Imperial duty, as the consultation itself was not the exercise of a power under the pension scheme (although it was about the proposed exercise of such a power).
The ways in which Warren J held IBM to have been in breach were (in summary): in certain respects, mis-leading information was given; in other respects, incomplete information was given (including not providing information reasonably requested by the pensions consultation committee); and not consulting with an open mind (in other words, it had already decided what it was going to do and it was not genuinely consulting about proposed action).
When IBM's consultation was considered in the context of the reasonable expectations that members had about ongoing pension provision, and also in light of both the legislative requirements about consulting on certain pensions issues and IBM's own "business conduct guidelines" and its "core values", the judge was less than impressed with the manner in which IBM had conducted the consultation process. He held that, in principle, it was possible for consultations that were not properly carried out to constitute breaches of the duty of good faith and that, indeed, was the case here.
Whilst the breaches of the consultation requirements here were less significant to IBM than might otherwise have been the case – because, in substance, Project Waltz was held to have been implemented in breach of IBM's Imperial duty, there was a bigger concern – the ruling that "defective consultation" can in principle constitute a breach of the duty of good faith might be significant in other cases.
This is because the sanction for breaching the requirements under the pensions consultation regulations12, where they apply, are limited – including a fine of up to £50,000 – and in the writer's experience usually do not cause employers to lose any sleep, if commercial exigencies mean that they are not fully satisfied. The prospect of being in breach of the duty of good faith might give employers more pause for thought.
It is not yet clear exactly what the consequences of such a breach in relation to a consultation exercise are. Further, the consequences will likely turn on the facts of any given case, but might include the employer having to repeat some or all of the consultation and suspending implementation (or more likely in practice, continued implementation) of its plans until that has been done.
Is this the end of the story for IBM?
Almost certainly not.
Firstly, the judge did not award any remedies as a consequence of the breaches that he identified. He would like to hear argument from counsel on the matter before doing so – therefore, the plan is for there to be a further hearing to that end.
It is important to note that the identified breaches do not necessarily mean that Project Waltz will be swept away and the closure to accrual (completely) undone. The "representative beneficiaries" were not arguing that IBM promised never to end defined benefit accrual – indeed, any such argument would have been bound to fail, in light of IBM's communication of the rights that it reserved – only that members had been given assurances (that are binding through the mechanism of the Imperial duty) that such accrual would not end for some unspecified period of time.
Also, it seems likely that there will be an appeal against elements of the judge's ruling by either or both of IBM and the trustee13.
What does the case mean for other employers?
Many employers have closed their defined benefit pension plans to future accrual over the past years and the case does not mean that all those exercises need to be reviewed and undone – far from it.
The facts of IBM were particularly extreme, including: a history of excellent defined benefit pension provision that was part of the company's culture, although both the provision and the culture changed over recent times, much to the annoyance and disappointment of many "IBMers"; binding assurances having been given about future pension provision and, as part of that, a failure to recognise – and therefore pay due regard to – reasonable expectations that IBM engendered in members; and a materially deficient consultation process.
Nevertheless, at the same time, it would be wrong for employers to think, "we're not IBM, this won’t happen to us". Other significant cases of recent times have triggered "copy-cat action", for example, the (not dissimilar) case of HR Trustees v German14 , which has resulted in many employers and trustees looking closely at situations where there has been a move from defined benefit to defined contribution pension provision at some point in the past.
Further, the issue as to compliance with the Imperial duty – and the good faith duty when consulting or entering into contractual arrangements about pension - is not confined to situations of ending defined benefit accrual. It applies, at least, to an employer's exercise of any discretionary power under a pension scheme.
Indeed, another aspect of IBM was a challenge to a proposed change (as part of Project Waltz) to a long-standing practice of IBM giving its consent to early retirements that enabled members to receive their pensions early, subject to a reduction that was less than "cost neutral". Instead, the future policy was to be that consent will be refused, other than in exceptional circumstances (meaning that members would have to resign and take a cost neutral early retirement pension from deferred member status, if they wished to retire early).
So, this is something that employers should take seriously, else – to quote from another of Irving Berlin's repertoire - they might find, like IBM with its ongoing litigation over three years after it started to dance, "The Song is Ended (but the Melody Lingers On)".