DEATH IN SERVICE BENEFITS Dropping while you work

Jamie Winter, Towers Watson, outlines his ideas for redesigning death in service benefits

In a nutshell: 
  • pensions legislation is starting to have a dramatic effect on the way in which death in service benefits are designed and delivered in the UK
  • despite changes in government and the law, as well as fluctuating economic cycles, these legislative pressures appear to be accelerating
  • now is the time for employers and trustees to consider how their death in service benefits stack up against these pressures.

The provision of group death in service benefits by employers and trustees in the UK used to be a straightforward affair; however, such simplicity was unlikely to last and it seems that three main “complications” have contributed to this in recent years.

In the UK, death in service benefits are inherently linked to pensions for tax and legislative purposes – and UK legislation is, of course, inherently linked to European Union law. In addition, we have not been able to avoid boom and bust economic cycles and we keep living much longer than anybody anticipated.

The last of these is a very good thing, but it does mess up the assumptions on which the design and cost of our death benefits have been based. So, as a result of these issues employers and trustees are rethinking their strategy for these benefits and are redesigning them to suit our modern age.



Added complexity


The 1989 Budget introduced the earnings cap, which restricted the amount of salary (£75,000 at the time) on which tax efficient pension contributions and benefits could be delivered. The “approved” pension regime could still provide valuable and tax efficient death in service benefits, but within certain limitations: a maximum salary of the earnings cap, a maximum lump sum death in service benefit of four times capped salary and a maximum dependant’s death in service pension (DDISP) of 4/9ths of capped salary.

DDISPs were taxable as income, but lump sums could be paid entirely tax free provided they were distributed through an appropriate discretionary trust. Also, the employer could claim tax relief at the corporation tax level on any insurance premium paid for approved death in service benefits, so provision and distribution were both highly tax efficient.

For those requiring benefits above the approved level, an “unapproved” market developed, which could provide similarly tax efficient benefits but without corporation tax relief. This approach essentially held good for the next 15 years. By 2006 the earnings cap had risen to £108,600, but it was not keeping pace with wage inflation and so more employees were caught by it. Also in the intervening period the nature of pension scheme designs began to change. Final salary schemes were becoming far less affordable as baby boomers began to retire and to live a financially sound and surprisingly long retirement. Defined contribution (DC) pensions therefore came to the fore and death in service benefits were provided alongside these, usually on an insured basis because the DC scheme had no facility to self finance them.

Then on 6 April 2006 – known as A Day – came the tax simplification legislation for pension schemes, which sought to sweep away the many existing regimes and replace them with one simple approach, principally consisting of two allowances: a tax efficient “input” amount for contributions called the annual allowance, which was set at £215,000 in 2006 and rose to £255,000 by April 2010; and, a tax-efficient “output” amount for benefits called the lifetime allowance (LTA), which was set at £1.5m initially and then rose to £1.8m by April 2010.

Although the headlines of the new approach were straightforward, complexity arose in trying to transition existing pension members to the new regime. Two forms of transitional protection were offered – primary and enhanced – and employees could apply for either or both of these depending on their individual circumstances. This in turn raised a number of issues in relation to the provision of group death in service benefits.

For example, it was no longer tax efficient to provide unapproved policies for lump sum death in service benefits, so a viable alternative was required for members with life assurance benefits in excess of the LTA. Also, the earnings cap disappeared, as did the DDISP maximum of 4/9ths of salary, so for many schemes their trust deed and rules had to be amended. Employees who took protection may have had to leave their current pension schemes, as a result of which they may suffer a reduction or loss of their death in service benefits. And, paying a premium for death in service benefits for a member who applied for protection could mean that the transitional protection is lost, so these members should be identified and assessed.

Some of these issues could be considered well before A Day, so a strategy could be put in place to deal with them. However, some were not clarified until very late in the day, such as the treatment of unapproved policies, which was only issued by HM Revenue & Customs in February 2006. As a result of this, unapproved policies had to be replaced either by “excepted group life” or by “single member relevant life” policies.

The LTA was deemed a generous value when compared against the previous earnings cap and at the same time lump sum death in service benefits were becoming cheaper due to improving mortality. Conversely, however, DDISPs were becoming more expensive because dependent beneficiaries were living longer and therefore claiming more benefit. For these reasons – combined with the fact that low investment returns were also making DDISPs more expensive – there began to be a shift away from providing DDISPs and into the provision of additional lump sums, which were tax free up to the LTA. One or two employers also introduced the option for their staff to decide annually whether to convert some, all or none of their DDISP into a lump sum alternative.

Further changes

The government’s recent review of the pensions taxation system has resulted in a further change which will affect death in service benefit designs: the LTA is scheduled to reduce from £1.8m to £1.5m from April 2012. This change is likely to bring with it a whole new raft of transitional protection for members with benefit values between £1.5m and £1.8m. We have not seen the detail of the protection arrangements at this stage, but we expect them to create another slice of complexity for employers and trustees to consider.

And it is not just tax legislation which is causing a reassessment of death in service designs. The government is also currently reviewing the default retirement age (DRA) – 65 in the UK – and has indicated its desire to remove it entirely. If this happens and there is no exception granted for employee benefits such as these, then suddenly employers will need to source death in service cover for employees who remain in service beyond the DRA. Whether such cover can be delivered in an affordable and sustainable way is still open to debate, but it certainly seems likely to prompt a further review of the death in service benefit levels in place.

Design elements

There are, of course, many ways to design a death in service benefit scheme, but if I was starting with a blank sheet of paper and wished to accommodate legislation as much as possible while controlling costs, then my scheme design would probably include the following elements:

  • Benefits provided through a standalone, “life assurance only” discretionary trust. This makes it easier to provide cover for employees who opt out of the pension scheme for any reason.
  • A core lump sum of 2x salary, or perhaps a fixed lump sum value. This ensures an affordable but still valuable level of cover.
  • An additional lump sum of, say, 4x salary, payable only where the employee has a qualifying dependant on his or her death. This avoids the increasing cost of DDISPs and should remain affordable into the future.
  • Maximum lump sum benefits may be capped at the LTA value or excepted policies may be established to provide any excess.
  • Members are included up to the age of 75, the maximum age insurable at the moment. This age limit may increase later where that is possible and/or necessary.

As with the A Day changes, this suggested design is simple to describe, but it is, of course, the transition from existing approaches which will add complexity. Nonetheless, the legislators’ ongoing tinkering with our pensions systems will no doubt continue to add weight to the redesign arguments for death in service benefits.

Jamie Winter is head of Towers Watson’s healthcare and risk consulting practice;


Jamie Winter is senior health and group benefits consultant at Towers Watson.