Revised Estimates of Compensation Costs, April 2004

by Dr. Ros Altmann

(All material on this page is subject to copyright and must not be reproduced without the author's permission.)

Paper to show calculation of costs of compensation for lost pensions


The evidence in this paper and my earlier research suggests that the cost of compensation will be very small, in the context of money spent on pensions as a whole. This is not a ‘black hole’ and based on the best available data, using a range of assumptions, the ongoing costs are easily affordable, if the schemes are run on and annuities are not bought. The savings to the taxpayer of not buying annuities could be significant.

This paper estimates that compensation would be unlikely to cost more than about £76 million a year (index linked) spread over 30 years. It could cost substantially less than this.

The paper uses new figures from OPRA’s Pension Schemes Registry, adjusting for preponderance of small schemes, with smaller pension entitlements and figures from the major independent firms of trustees who are involved in statutory winding up of final salary schemes.

To put the figures in context, we spend £14 billion a year on tax relief for pension contributions to non-state pensions. £7billion a year of this goes to the top 10% of taxpayers and about £4billion of this goes to the top 2.5% of taxpayers every year. This only goes to those contributing to pensions.

Furthermore, the sum of around £100million a year is already available in the DWP budget for each of the next three years, without allocating any new funds, since there is a sum of this magnitude in the accounts ‘unallocated’ at the moment. We could earmark it for this purpose, without any added strain on the public finances.

Finally, the Chancellor’s recent announcement of pension simplification includes measures to allow the very wealthy to accumulate up to £1.5million, rising to £1.8million in pensions over their lifetime and also allows them to take higher tax free lump sums out of their pensions. Estimates produced by the Treasury, show that the cost of allowing these top earners to take higher tax free lump sums is likely to be £200million over the next three years, and more after that.

This additional spending of hundreds of millions of pounds will only go to those who have their own occupational or personal pensions, yet all taxpayers have to pay for this, whether they have their own pensions or not. The rationale that taxpayers should not fund compensation is partly based on the argument that all taxpayers should not have to assist people who have lost money in private pensions, because not all taxpayers have occupational or personal pensions themselves. This argument seems hollow, in the face of billions of pounds in tax relief for those who contribute to pensions and the Chancellor’s plans to spend an extra £200m of taxpayers’ money giving top earners more tax-free cash.

Given the public interest, moral and legal arguments in favour of restoring these people’s pensions, it seems difficult to justify failing to settle this matter.


I have calculated the potential costs of compensation for members of final salary schemes winding up since 1995. The results from these new calculations suggest a lower figure than my original work, which was designed to be an upper bound figure, based on extremely pessimistic assumptions. My original study suggested a possible maximum cost averaging £93million a year for 50-60 years, which would be over £5billion. The new figures are coming in closer to £2.3 billion, which could be an average of £76 million a year or less, over only 30 years

Data from the independent trustee firms, involved in the largest schemes, suggest compensation would cost an additional £1.265billion and assuming 500 smaller schemes, the extra required could be around £1billion for these, giving a total of £2.265billion.


Using latest figures from OPRA, total assets from independent trustees plus estimate of small insured schemes

Trustee data:

I have now had at least some information from almost all the major firms of independent trustees. These are the main firms involved in wind-ups and the total size of the assets for schemes they have been winding up since 1995 or 1997 is reported to be around £2.3billion.

Analysing data and talking to partners at leading actuarial firms I have ascertained that a good rule of thumb for trying to assess how much more these schemes in wind up might need, if they were to be able to meet their pension liabilities if the scheme is running on an ongoing basis, would be to add an extra 50% to the assets. This assumes they were fully funded on an MFR basis. We know that most schemes were underfunded on the MFR at time of wind-up, so we assume that an extra 10% will be needed to cover this shortfall. On this basis, the assets of these winding up schemes would need to be increased by £1.265 billion, to be able to pay the pensions on a continuing basis.

OPRA figures:

I have just received new figures from OPRA (1), which give number of schemes winding up as at March 2004. OPRA have carried out this analysis for me, to try and help us identify more accurate data. Unfortunately, the OPRA database was not designed to track details of schemes and is very difficult to manipulate, however the figures show that, as at March 2004, the number of final salary schemes which were winding up was 1234. The number of members in these schemes was 174,228. Many are small schemes (72% have less than 100 members and 15% have 11 members or less). This is obviously way above the 60,000 figure from Andrew Smith. It could be that the DWP figure excludes many very small schemes. It may also be that the 60,000 figure refers to members who have lost some entitlements. Pensioners will mostly have had their pensions paid in full, whereas non-retired members have lost out, so members of schemes would number well over 60,000 but only 60,000 may have suffered losses.

Small insured schemes:

Since the independent trustees cannot account for more than a few hundred schemes and OPRA has over 1200 schemes in wind up, how do we account for the ‘missing’ schemes? Of course, some of the schemes in the OPRA database will be wound up without a deficit. In addition, it is likely that the balance of the schemes in wind up are smaller insurance company schemes, which have not been passed to these independent trustee firms on winding up. Assuming that, say, 600 schemes are small insured arrangements, I need to identify how much extra money would be needed to make up shortfalls in these schemes.

I have spoken to some trustees who are involved in winding up small insured schemes (which do not generally reach the major independent trustees, since the insurance companies often use different trustees, such as those that I have spoken to). These trustees, some of whom specialise in winding up these small insured schemes, tell me that the average asset size per member is very small – often £10,000 - £20,000 or so, which would obviously give a very small amount of pension entitlement - and the Government Actuary’s Department table from 2001 (supplied by Mercer) shows an average asset size per member of £17,000 (2), which is consistent with this.

If I assume that these 500 schemes have an average of 100 members and the average shortfall of pension is, say, £10,000, this will add a further £1billion to the cost of compensating for lost pensions. The total will then be just around £2.265billion.

The compensation required would be around £76 million a year (index linked) for 30 years.


(1) Data supplied on 2nd April 2004 by OPRA. Showing schemes in wind up at March 2004, with number of schemes and number of members, broken down by scheme type and number of members.

(2) GAD 2001 Survey, reproduced by Mercers show that schemes winding up or frozen, with 12-100 and 100-1000 members have an average assets size per member of £17,000 and these schemes are smaller than those not in wind up. Average assets of winding up schemes with 100-1000 members are £3.3million, whereas for live schemes the average assets total £11million. For schemes with 12-100 members, the comparable figures are £0.7million average assets if winding up and £1.1million for live schemes.

© Dr. Ros Altmann  |  Home  |  Profile  |  Disclaimer