Iain Clacher and Con Keating
Since their heyday in the 1990s, UK occupational DB schemes have been winding up at a rate of about one every two days. Their replacement, DC schemes, are a poor substitute; they are tax-advantaged savings schemes rather than pensions. They fail to satisfy the desire for a specified income in retirement, which has been found repeatedly in numerous polls and surveys of employees. The members of a DC scheme bear all its risks; a task for which most are ill-equipped. DC pensions also look like they will produce highly variable retirement outcomes, many of which will be grossly inadequate, offering little comfort or certainty even as retirement approaches[i].
Of the 5,400 schemes[ii] still covered by the Pension Protection Fund only a small minority are still open to new members and future accrual. DB pensions are increasingly confined to public sector employees.
There is a widely held belief that DB pensions are unaffordable and prohibitively expensive. It is undoubtedly true that the cost of providing DB pensions has risen substantially in recent decades, giving rise to the paradox that the pensions being offered have themselves not increased to a similar degree as the proportion of new capital which has been invested. In this, and subsequent articles, we shall investigate the role that legislation and regulation have played in this transformation.
The term ‘defined benefit’ (DB) is a quite recent introduction, imported from American parlance and was not used by the Regulatory Authorities. In the 1950s, 1960s, and until the late 1970s, they were simply pensions, or occasionally more specifically, as final salary pensions or superannuation. The name ‘defined benefit’ has also been a source of some confusion; it was just shorthand for schemes where the benefit ambition, the pension, an income in retirement, was formulaically described or defined. More recently it has been adapted to distinguish between schemes which have explicit or implicit guarantees from those where there are no guarantees, and the fulfilment of the ambition depends upon the performance of its pension fund[iii].
In the early post-war period, occupational pensions carried, for beneficiaries, a certain air of privilege; the status of relatively menial, low paid jobs could be enhanced by their description as pensionable positions. As pension provision grew to be widespread, this privilege steadily dissipated and by the mid/late 1970s, occupational pensions had become almost a standard term of employment; particularly so where the workforce was large and unionised.
Baldly stated, the pension element of the employment contract was:
‘I, as employer, agree with you, as employee, to enrol you in my (DB) pension scheme and to contribute each year the cost of one year’s accrual estimated at that time for that year.
In the fat years, my contributions may be reduced by valuation surplus. In the lean years, my contributions may, if I so choose, be increased. But I always retain the right to terminate payments to the pension fund (with no obligation to make up any deficit and no obligation to continue future accrual).’
There were no guarantees here and no enforceable debt on the employer. There is no liability or shortfall to be included in the sponsor employer’s balance sheet. In this framework, it is hardly surprising that the primary purpose of the fund should be regarded as being to pay the pensions when due.
Becoming an ever-more binding obligation
It is possible for reasonable expectations arising from behaviour, e.g. from prior practice, to convert into legally binding terms of the pension agreement part of the contract of employment. But, normally, the employer communications were boiler-plated to prevent that happening and case law evidencing the conversion of expectations into legal rights, in general, has come out in favour of the employer[iv].
If we consider the benefit ultimately payable, clearly the investment accumulation on the contribution through time is far larger and more important than the initial contribution, and the sponsor is not involved in the investment accumulation. However, when we introduce legislation making it a debt on the employer, we are changing that. The employer is now guaranteeing the investment return, and the role of the fund becomes defraying the employer cost of that investment return and serving as collateral security for the accrued benefits.
Regulation over the 1960s and 1970s was light. The scheme’s trust deed governed its management and operations. Schemes could and did make hard promises, which might be described as guarantees; for example, some offered fixed indexation of pensions in payment. Usually and almost invariably the directors and other officers of the sponsor were members of the scheme, with some also trustees. With pensions then a more important part of their total compensation than now, they were anxious that it did not fail. This community of interest now no longer applies but was previously critical in ensuring good governance through common interest. The chairman of the first scheme joined in the 1960s by one of the authors (Keating) would have been delighted to be described as a gentleman amateur in the best British tradition.
In recent decades, there has been a radical shift in emphasis in the governance and trustee duties, away from this quasi-paternalistic model, to one in which employee representation and professionalism dominate. Professional trustees have been described as ”regulators in residence” while the lay trustees for their part are being slowly drowned in torrents of regulatory guidance
In this regard the extensive emphasis in recent consultations on propriety in ’fit and proper’ tests for trustees gives us cause for concern. The financial soundness condition is reminiscent of the Victorian idea that only men of property might vote. With propriety defined as “conformity to conventionally accepted standards of behaviour”, innovation and inventiveness seem unwanted, which is concerning as diversity, and more specifically cognitive diversity, is frequently viewed as highly desirable for problem solving and for effective decision-making.
Following the introduction of the PPF, practices when sponsor employers were in difficulty or insolvent changed. However, in the early period, of course, some companies with DB schemes did fail, but failure was anticipated. The trust deeds invariably contained a specified way of re-allocating benefits, even if that may only have been at the discretion of the trustees. They were taught to cut benefits more for younger, working members (including former employee, deferred members) than retired members, since the former had time to rebuild their rights with another employer. Trustees had the right and in fact were expected to run the scheme on (in wind-up) rather than buy annuities, to lower and spread the costs over time.
The first externally imposed and determined guarantees came, in 1978, with the state second pension, SERPs[v]. National Insurance contributions were reduced (rebated) for those contracting out, but the scheme had to offer benefits at least as good as those of the state scheme – the infamous guaranteed minimum pensions (GMPs). These were not considered onerous, and there was a flurry of new contracted out schemes (COPS) created. As was predicted by some, the terms of GMPs did become more onerous: notably with the introduction of indexation from 1988. GMPs are the longest-running open sore of the pensions industry; their administrative cost has often far exceeded the value of the rebate received. Painful though these legacy issues have been, with the ending of contracting out in 2016, they are not material in the ongoing question of DB pension affordability. The introduction, in 1985, of compulsory revaluation of deferred member benefits is far more important. This was compounded, in 1986, by the changes to cash equivalent transfer values.
In the 1980s, the issue was not scheme deficits but the magnitude of scheme surpluses. The valuation practices at this time were questionable, smoothing of assets[vi] and liabilities when determining contribution rates. In the late 1980s, schemes which were more than 105% funded on a prescribed basis were required to remedy the situation over five years in one of three ways. The route most frequently taken was to increase benefits, but more than a few opted to take refunds in the form of reduced contributions as well. The least popular of the three options was payment of surplus to corporate sponsors, which was taxed at 35%. This regime undoubtedly contributed to the willingness of companies to use scheme surpluses in the redundancy packages offered in their restructurings in response to the 1991-1992 recession. Early retirement on full pensions figured large. However, the impact of this has often been overstated. There was only one year in the 1990s in which DB pension contributions declined[vii], and then only very marginally; the aggregate addition to liabilities was perhaps 10%.
The various additions to pension costs over these decades added to the benefits payable; they do not however, explain the current paradox of liability values having grown far faster than the actual benefits promised. The changes we have described added perhaps 15% – 20% in total to scheme liabilities, but schemes remained open. There was no systematic movement to close to new members or future accrual, and few schemes were voluntarily wound up. The elephant in the room is that none of this addressed the fundamental questions of member security. The positive is that these pension schemes emerged unscathed from a radically changing economy, rampant inflation, and several severe financial disruptions.
The Pensions Law Review, the Goode Report, is a convenient marker for the end of this period in 1993, as what comes next is the modern era of pensions regulation.
[i] A fuller discussion of DB versus DC is available from the authors on request. This takes the form of slides and notes to a presentation at the Royal Society in early 2019. (or from firstname.lastname@example.org)
[ii] Down from 7800 in 2005.
[iii] Collective Defined Contribution (CDC) In this and subsequent articles we shall use the term to describe occupational schemes authorised by the Inland Revenue / HMRC following the 1921 Finance Act model. We will not cover unauthorised schemes, the pension policies issued by insurance companies or money purchase (Defined Contribution(DC)) schemes.
[iv] See,for example: http://www.bailii.org/ew/cases/EWHC/Ch/2011/960.html
[v] State Earnings Related Pension
[vi] In smoothing asset values the UK appears to have been almost unique internationally
[vii] See ONS MQ5