Rt Hon Lord Lilley



    The Foundation?s main activity is to commission and publish original papers by

    independent academic and other experts on key topics in the economic and

    social fields, with a view to stimulating public discussion on the performance

    of markets and the social framework within which they operate.

    The Foundation is a registered charity and a company limited by guarantee.

    It is independent of any political party or group and is financed by the sales of

    publications and by voluntary donations from individuals, organisations and


    The views expressed in publications are those of the authors and do not

    represent a corporate opinion of the Foundation.


    David Lipsey (Lord Lipsey of Tooting Bec)

    Members of the Board

    Viscount Chandos

    Gavyn Davies

    David Edmonds

    John McFadden

    Baroness Noakes

    Brian Pomeroy


    Philip Collins

    Deputy Director

    Beth Breeze

    SMF Save Our Pensions Text 16/9/03 12:06 am Page 1

    First published by The Social Market Foundation, 2003

    The Social Market Foundation

    11 Tufton Street

    London SW1P 3QB

    Copyright ? The Social Market Foundation, 2003

    The moral right of the authors has been asserted.

    All rights reserved. Without limiting the rights under copyright reserved above,

    no part of this publication may be reproduced, stored or introduced into a

    retrieval system, or transmitted, in any form or by any means (electronic,

    mechanical, photocopying, recording, or otherwise), without the prior written

    permission of both the copyright owner and the publisher of this book.

    ISBN 1-874097-53-4

    Cover design and typesetting by Beaufort 5

    SMF Save Our Pensions Text 16/9/03 12:06 am Page 2


    Peter Lilley


    September 2003

    SMF Save Our Pensions Text 16/9/03 12:06 am Page 3


    I am very grateful to all those who have so kindly commented on successive

    drafts of this paper particularly Professor David Blake of The Pensions Institute,

    Carl Emmerson of the Institute of Fiscal Studies, Mary Francis of the ABI,

    Lorraine Fraser of Winterthur Life UK, David Harris of Watson Wyatt, John Jory

    of the Building & Civil Engineering Benefit Schemes, Tim Keogh of William

    Mercer Ltd, Andrew Mitchell MP, Ann Robinson of ASSP, Mark Rowlands of

    Winterthur Life UK, Andrew Tyrie MP, Mike Wadsworth of Watson Wyatt, David

    Willetts MP and Andrew Williams. My thanks are also due to the DWP and

    Government Actuary?s Department for providing factual and quantitative


    None of them are responsible for its errors or omissions nor necessarily

    endorses its analysis and conclusions.

    SMF Save Our Pensions Text 16/9/03 12:06 am Page 4


    Glossary of Terms 6

    Executive Summary 7

    Introduction 14

    Compulsion: How Much Pension Provision? 16

    Funding: Save Now or Tax Later? 24

    Pension Age: How Long to Work? 37

    How to Cut Pension Costs 44

    Conclusion 49

    Annex A How the UK State Pension System Works 51

    Annex B Basic Pension Plus 55

    Annex C Australia?s Experience of Compulsory Personal Pensions 56

    SMF Save Our Pensions Text 16/9/03 12:06 am Page 5



    (See Annex A for a guide to the complexities of the British pension system)

    ABI Association of British Insurers

    APP Approved Personal Pension

    ASP Additional State Pension (Pension additional to BSP, was SERPS now S2P)

    BPP Basic Pension Plus

    BSP Basic State Pension

    DWP Department for Work and Pensions (formerly DSS)

    DSS Department of Social Security

    GAD Government Actuary?s Department

    HET Higher Earnings Threshold (S2P accrues at 10% of earnings between LET and

    HET and at 20% between HET and UEL)

    LEL Lower Earnings Limit (Level of earnings – set roughly equal to BSP -below which

    employees neither pay NI nor accrue any entitlement to BSP and S2P)

    LET Lower Earnings Threshold (Set roughly equal to half national average earnings.

    Employees earning between LEL and LET accrue a flat rate element of S2P as if

    they earned LET)

    MFSP Mandatory Funded Second Pension (The pension generated by NI rebates

    invested in a fund to replace the unfunded S2P)

    MIG Minimum Income Guarantee (Level to which retirement incomes will be topped

    up by Guarantee Credit)

    NI National Insurance

    NIC National Insurance Contribution

    NI Rebate Element of employee?s and employer?s NICs paid into private pension fund of

    those who contract out of ASP

    ONS Office of National Statistics

    Pay-As-You-Go Scheme in which current taxes or NI contributions are used to pay pensions of

    people already retired rather than invested to pay future pension liabilities.

    PC Pension Credit (composed of two elements: Guarantee Credit which tops up

    retirement incomes to the MIG level and Savings Credit which partly

    compensates for impact of means testing on income from private saving)

    SERPS State Earnings Related Pension Scheme

    SPA State Pension Age

    S2P State Second Pension

    UEL Upper Earnings Limit (Level beyond which higher earnings accrue no extra S2P)

    SMF Save Our Pensions Text 16/9/03 12:06 am Page 6


    Reform of Britain?s pension system must address four key questions.

    They are:

    Compulsion how much should people be required to provide for

    their old age?

    Funding how much of that compulsory provision should be funded by

    genuine savings and how much left to future tax payers (by the

    state financing on a pay-as-you-go basis)?

    Pension age should people be required or encouraged to work longer before

    drawing their pension?

    Costs how can the portion of savings absorbed by the costs of

    running pensions and annuities be reduced?

    Compulsion: How Much Pension Provision?

    ? Compulsion is undesirable. But it is unavoidable. Any civilised society is

    bound to provide at least a means tested safety net for those who reach old

    age without providing for their retirement. That creates disincentives to

    save which cannot be eliminated, only spread more thinly further up the

    income scale. It also means that the prudent, who voluntarily provide for

    their own retirement, are compelled through the tax system to pay for

    those who improvidently fail to do so. It is better to require everyone who

    can do so during their working lives to make provision for their own

    retirement. That will not be an extra burden on those who would save

    anyway. Indeed, they will then not be compelled to pay so much tax to

    support the improvident in the future.

    ? We already have a compulsory second pension. All employees are required

    to pay via their National Insurance Contributions either into the State

    Second Pension (previously SERPS) or into an approved occupational or

    personal pension scheme offering similar or superior benefits. The key issue

    is whether the level of this compulsory provision is sufficient.

    ? The level of compulsory provision should be sufficient to lift people clear of

    the means tested safety net. Means testing discourages people on low

    incomes from saving and it provokes resentment among those who


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 7

    nonetheless do save yet are little or no better off as a result. In Australia

    making everyone in work provide adequately for their retirement has been

    widely welcomed especially by younger people.

    ? Coverage of the compulsory second pension should be extended to the

    self- employed.

    ? The government is committed to raising the Minimum Income Guarantee in

    line with earnings during this parliament and projects it forward indefinitely

    on that basis. Yet state pensions are set to rise only in line with prices. If this

    continues the MIG is bound to overtake the level of state and compulsory

    pensions. To prevent this happening, the Basic State Pension and the MIG

    should be statutorily indexed to prices and, when resources permit, any real

    increases made by raising both the BSP and the MIG by the same cash amount.

    ? Once an adequate Second Pension is in place it will be possible to phase out

    the Savings Credit system whose complexities and withdrawal rates act as

    disincentives to save over a broad swathe of incomes.

    Funding: Save Now or Tax Later?

    ? The UK has more investments to meet future pension requirements than

    the rest of the EU put together. Nonetheless these still only cover 40% of

    our future pension liabilities. The government has set a laudable target of

    raising that proportion to 60% by the middle of the century.

    ? This will only be achievable if more people contract out of the unfunded

    State Second Pension scheme into genuinely funded pensions.

    ? Sadly Stakeholder Pensions have had negligible success in persuading more

    people to contract out of the State Second Pension.

    ? Those who contract out receive a rebate from their National Insurance

    Contributions payable directly into their personal or occupational pension

    fund. The level of the rebate is set on the basis of the Government Actuary?s

    estimates of the amount needed to fund a private pension broadly

    sufficient to replace the pension entitlement that employees would have

    accrued in the State system.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 8

    ? But contracting out of the unfunded Additional State Pension system into

    a personal pension has never been attractive for those on low earnings. This

    is because the lower a person?s earnings the smaller their rebate. Charges

    do not decline proportionately since there are fixed costs in setting up and

    running a fund. So charges absorb a disproportionate share of the small

    rebates received by those on low earnings.

    ? Consequently the option of remaining in the unfunded State Additional

    Pension system has been retained for the benefit of those on low and

    intermittent earnings.

    ? That problem was inevitable under SERPS since the accrual of pension and

    the corresponding rebate for contracting out were directly proportional to

    earnings. However, under S2P, employees with earnings below roughly half

    the national average accrue a flat rate entitlement to pension. Yet if they

    contract out they still receive a rebate proportionate to their earnings so it

    remains unattractive for them to contract out.

    ? Those on low earnings should be entitled to a flat rate rebate equivalent

    to the full value of the pension entitlement they accrue at present under

    S2P. That would be sufficient to justify the fixed costs of setting up a

    personal pension. It would then be possible gradually to replace the

    unfunded pay-as-you-go S2P by a Mandatory Funded Second Pension of

    equivalent value. Those on earnings above the Lower Earnings Threshold

    would receive a rebate comprising both the flat rate element for earnings

    up to the LET and an earnings related element for earnings above the LET

    as at present.

    ? This proposal would come in progressively with each cohort of young

    people entering employment. I envisage that everyone in work who was

    born, say, thirty years or less before the new system starts would be

    required to have a Funded Second Pension. They would receive a NI rebate

    equivalent to the value of the old S2P accrual which it would replace.

    That payable into their own personal or occupational fund. Those born

    more than 30 years before the start date would retain the choice, as at

    present, of remaining in the unfunded S2P or opting into a funded

    private pension.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 9

    ? The government originally envisaged compelling some employees to have a

    private Funded Second Pension to replace the unfunded S2P. But only

    those with earnings above the Lower Earnings Threshold would have been

    required to contract out in this way. It now envisages that all those with

    earnings below the LET will lose the right to contract out of the S2P. Those

    above this level will be encouraged to contract out. If they do they will

    receive a rebate equivalent to the flat rate and earnings related elements of

    S2P but if not they will only accrue the flat rate of S2P. This would create

    an unfair two class system and difficulties for those whose incomes

    fluctuate above and below this level.

    ? My proposal would enable the government to meet its target of 60%

    privately funded pensions by mid-century; enormously boost committed

    long-term savings and investment; and dramatically reduce the long-term

    tax burden. More important, over time it would ensure that everyone had

    their own pension fund – the biggest extension of capital ownership since

    the growth of home ownership.

    ? Once people own a personal pension fund they can readily choose to save

    more than the minimum ? and Australian experience suggests many will do

    so. Ownership of a pension fund will also give everyone a direct stake in

    national prosperity.

    ? The government should guarantee that if for any reason the element of

    anyone?s Funded Second Pension derived from the flat rate rebates is not

    sufficient at retirement age to buy an annuity equivalent to the flat rate

    element of S2P the state will top up the annuity payments to that level.

    Earnings related rebates and voluntary additional contributions will not be

    covered by this guarantee.

    ? To prevent fund managers betting against the guarantee if the fund has

    underperformed they would be required to invest on a Prudent Person

    basis. These rules could be enshrined in statute as in Australia. Also fund

    managers would be required to invest the guaranteed and non-guaranteed

    elements of the funds on the same basis.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 10

    Pension Age: How Long to Work?

    ? Unless it moves to a Mandatory Funded Second Pension, the government

    could only reach its target of 60% funded pensions by 2050 by raising the

    State Pension Age to make people work longer.

    ? Given that the life expectancy of those aged 65 has increased by roughly one

    month every year for several decades there are strong pressures to raise the

    SPA annually by that amount simply to contain the cost of state pensions.

    ? As far as possible the government should avoid telling people when to

    retire or forcing them to work longer by raising the SPA. The most it should

    do is to raise, in line with average life expectancy, the age at which people

    can draw their tax free lump sum other than to buy an annuity.

    ? The great advantage of everyone having their own pension pot is that with

    ownership comes choice. People will be free to choose when to retire once

    their fund can provide an income sufficient to keep them from dependence

    on means tested benefits. At the same time ownership of a personal

    pension fund will give them a double incentive to work beyond the normal

    pension age ? to save more for fewer years of retirement.

    ? Compelling people to work longer would bear hardest on those who work

    in stressful, manual and lower paid jobs. These are the very people whom

    the present pension system treats most unfairly. They pay in on the same

    terms as everyone else. Yet they have shorter life expectancy and so

    typically draw their pensions for fewer years. The less well paid subsidise

    the retirement of the higher paid and longer lived.

    ? Government should encourage providers to remedy this by relating

    pensions and annuities to the life expectancy of different income groups.

    This would mean that the lower people?s earnings during their working life

    the less they would pay to buy a given annuity income. So the less well paid

    would not need to accumulate so large a pension fund before they can

    choose whether to retire or continue working (possibly part-time).

    ? To enable annuity providers to do this the government could make

    available, in a form equivalent to their tax code, a summary of each


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 11

    person?s life time earnings from the National Insurance Recording System.

    The likely growth of affinity group providers once everyone has their own

    pension pot will also lead to provision of annuities more fairly reflecting

    the life expectancy of lower income groups.

    How to Cut Pension Costs

    ? The costs of running pension funds typically reduce the value of the final

    pension pot by over a third. Over half of this is the cost of acquiring


    ? Requiring everyone to have a pension fund would be the single most

    effective way of reducing those costs. Compulsion would eliminate costs of

    persuasion, largely remove the need for advice, simplify and standardise the

    product, and spread overheads more widely. So costs might be reduced by

    as much as a half, increasing the final pension by as much as 13%.

    ? Insurance is a good way to pool risks which are stable across a population.

    But there is no way to hedge the risk that average life expectancy might

    suddenly outstrip past trends ? e.g. because of amagic bullet cure for cancer.

    ? Annuity providers have to make provisions against possible but unlikely

    developments in longevity. If those developments fail to occur and

    mortality is in line with central estimates pensioners will collectively have

    paid for more years of retirement than they receive. On the other hand if

    annuity providers fail to make adequate provision for future increases in

    longevity they may prove unable to pay pensions they have promised. The

    state would have to step in.

    ? Most of the uncertainty annuity providers must provide against concerns

    the number of people who may live well beyond current average life

    expectancy. So if the Mandatory Funded Second Pension was only required

    to cover the first twenty years of retirement and the State provided an

    unfunded S2P for the years beyond age 85 the state would be bearing

    much of the unquantifiable longevity risk.

    ? The cost of twenty year fixed term annuities should be disproportionately

    cheaper than full life annuities. Anyone wanting a retirement income


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 12

    above the Second Pension level would still have to buy an additional full

    life annuity


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 13


    The biggest domestic problem facing every developed country is how to

    finance pensions as people live longer and have fewer children.

    Britain is better placed to cope with this than most of our European

    partners. They rely almost entirely on taxation to pay for pensions. So the

    growing burden of their aging populations will mean rising tax burdens on

    declining work forces. By contrast, we have encouraged people to contract out

    of the unfunded Additional State Pension and save in occupational and

    personal pension schemes. As a result we have built up more investments to

    meet future pension liabilities than all the rest of the EU put together.

    Nonetheless, UK pension funds only account for 40% of future pension

    liabilities ? the other 60% falling on future tax payers. The Labour government

    set a target to reverse those proportions so that 60% of pensions would come

    from savings by the middle of this century. That is an admirable target that a

    future Conservative government would, I hope, endorse.

    Unfortunately, having set itself that target, the government has actually

    moved in the opposite direction. It has increased unfunded state pension

    promises enormously while imposing a ?5 billion per annum tax on pension

    funds and creating disincentives to save by its Minimum Income Guarantee

    and Pension Credits.

    At the same time private pensions have been plunged into a crisis. Defined

    benefit schemes are closing to new employees at an accelerating rate.

    Employers are changing from defined benefit to money purchase and often

    contracting their employees en bloc back into the state system. Insurers are

    urging many employees to opt back into the unfunded State Second Pension.

    The stock market fall has provoked a crisis at Equitable Life, caused other withprofits

    pension providers to cut their bonuses and hit even harder those

    invested in unsmoothed funds.

    In the face of all this some commentators advocate retreat: abandon the

    60:40 savings target, leave it to tomorrow?s taxpayers to pay for our pensions

    and/or force people to work longer before they receive their state pension.

    To retreat from funded pensions would be folly. Britain?s funded pension

    provision is an enormous asset. We should see current problems as

    opportunities to strengthen it and build on it. But that needs new, imaginative

    and radical thinking. Sadly the government?s Green Paper, though it contained


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 14

    some useful minor reforms, did not remotely match up to the scale of the

    pension crisis. It largely sidestepped the four key issues facing our pension

    system which are:

    compulsion ? how much people should be required to provide for their

    old age;

    funding ? the extent to which compulsory provision should be funded by

    genuine savings or left to future tax payers;

    pension age ? whether people should be encouraged or required to work

    longer before drawing their pension;

    costs ? how the cost of pensions and annuities can be reduced.

    These are the issues addressed in the four sections of this report.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 15



    Should the state compel people to make adequate provision for their

    retirement? And if so, how much?

    In this chapter I do not address the issue of the extent to which any such

    compulsory provision should be financed by the state on a pay-as-you-go basis

    and how much by private savings. That is the question considered in the next


    The importance of the issue of compulsion has been accentuated by the

    erosion of incentives to save (through the extension of means testing to the

    majority of future pensioners) and the limiting of any reward for persuading

    people to save extra (through the cap on charges for stakeholder pensions). Yet

    the government?s Green Paper largely sidestepped it ? instead setting up a

    Commission to monitor whether voluntary provision is sufficient.

    Concerns about the adequacy of people?s provision for retirement have been

    growing. The Association of British Insurers claims there is a “pension shortfall

    of ?27 billion”.1 As the trade organisation for the providers of personal pensions,

    they have a vested interest in talking up the need for more pension provision.

    Nonetheless, since that figure was calculated the stock market has fallen

    much further, more pension funds have closed and Equitable Life?s problems

    have cast their shadow over other with-profits providers.

    As a result, calls for the state to compel people to make greater provision

    for their retirement have multiplied. Often such calls take the form of demands

    that the state require people (or their employers) to provide for a “second


    A surprising number of commentators seem unaware that all employees are

    already compelled to provide for a pension in addition to the basic state


    Ever since 1978 all employees have been required to make payments via

    their national insurance contributions either into the Additional State Pension

    (called the State Earnings Related Pension System until April 2002 when it was

    restructured and renamed the State Second Pension) ? or into an approved

    occupational or personal pension scheme offering equivalent or superior



    SMF Save Our Pensions Text 16/9/03 12:06 am Page 16


    People may be unaware of this because they never see the contributions

    they are making – since they are deducted at source from their pay and go

    direct to the National Insurance Fund or onwards to their pension fund.

    Is the State Second Pension adequate?

    The issue is not, therefore, whether a compulsory second pension needs to be

    introduced but whether the new State Second Pension or the rebates for

    opting into a private scheme will be adequate.

    At first sight the amount of provision people should make for their

    retirement is a matter for them. Why should the State compel anyone to make

    a minimum level of provision for themselves? I am on the libertarian end of the

    political spectrum. So I started from a presumption that compulsion should be

    avoided if at all possible. But I concluded that if, as any civilised society must,

    we make provision for those who cannot provide for their own retirement, an

    element of compulsion is inevitable. Either we compel all those who can do so

    to provide sufficiently for themselves; or the prudent, who voluntarily provide

    for themselves, will be compelled also to support via the tax system the

    imprudent who fail to do so.

    Virtually everyone would accept that, at very least, we must help those who

    simply could not afford to save for retirement whether through sickness,

    disability, caring responsibilities, unemployment or very low income when they

    were of working age. Not everyone would be so willing to help those who

    could have made provision for retirement but failed to do so. In practice it

    would be difficult to distinguish between those who would have saved but

    couldn?t and those who could have saved but didn?t. In any case, however

    much we may disapprove of the improvidence of the latter, they cannot be left


    So, at the very least, any decent society must have some means tested

    safety net for those who reach old age with insufficient provision for their

    retirement. However, once the state effectively guarantees a means tested

    minimum income in old age it inevitably creates two problems. It discourages

    people on low incomes from saving and it arouses valid resentment among

    those who do nonetheless save yet are little or no better off as a result.

    The only logical way to remove both disincentive and resentment is to

    require people of working age who can afford to do so, to provide for a pension

    at least equal to the minimum retirement income guaranteed by the state. Those

    who might otherwise have been improvident would then be forced to make

    SMF Save Our Pensions Text 16/9/03 12:06 am Page 17

    provision during their working lives. So they would no longer arouse resentment

    among the prudent nor burden the taxpayer. Those who would have made

    adequate provision for themselves anyway will not be directly affected by

    compulsion. Indeed, they will not be compelled to pay so much in taxation to

    support the imprudent in future. Those who cannot make such provision in any

    year (e.g. because of caring responsibilities) would be given credits towards such

    a level of retirement income. Ideally those credits would score as accruals for

    pay-as-you-go funds and be paid in cash for funded pensions.

    This was roughly the structure envisaged by Beveridge when he designed

    the welfare state. People were initially required to contribute during their

    working lives towards a basic state pension that was then set somewhat higher

    than the minimum level of income guaranteed by National Assistance. Over

    time, however, the minimum means tested income for the elderly (provided

    first by National Assistance and latterly by Income Support) has been increased

    to a level above the Basic State Pension.

    So long as the gap between the two was small comparatively few people

    suffered means testing of their savings. This government has greatly

    exacerbated the problem by setting the Minimum Income Guarantee nearly

    one-third higher than the Basic State Pension. This has substantially increased

    the number of people facing disincentives to save and feeling resentment if

    they do. The Minimum Income Guarantee in 2003/4 is ?102.10 pw against the

    Basic State Pension of ?77.45 pw for a single person.

    So someone with just the BSP of ?77.45 will be entitled to claim Guarantee

    Credit of ?24.65 to bring their total income up to the MIG of ?102.10. A

    pensioner with a small private pension of, say, ?10 pw on top of their BSP will

    currently get ?10 less Guarantee Credit ? just ?14.65 ? to top them up to the

    MIG level.

    The MIG on its own thus renders voluntary saving towards a modest second

    pension of up to ?24.65 pw completely pointless. Even a private pension

    somewhat higher than this represents a very poor net return for saving. For

    example, a pension of ?34.65 on top of the BSP giving a total income of

    ?112.10 leaves the pensioner only ?10 pw better off than if they had not saved

    a penny. Most occupational pensions are quite modest. About half are less than

    ?40 pw. They will barely lift pensioners above the level they could have got

    relying solely on the MIG.

    Moreover, the government has promised that the Minimum Income

    Guarantee will rise in line with earnings. Yet the Basic State Pension is still only


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 18

    indexed to prices. So the gap between the two is set to widen making the

    problems of disincentives and resentment progressively worse. Even if someone

    has a pension sufficient to lift them clear of the means tested zone at the

    point of retirement they may well find themselves subject to means testing in

    the course of their retirement as the MIG increases in line with earnings and

    overtakes their pension which is indexed to prices.

    The government hopes to mitigate the disincentive effect of means testing

    by introducing the Pensions Credit. This comes in from October 2003.

    The effect of this is to restore some, but not all, of the reward for saving to

    those who previously would have done just as well to rely on the MIG.

    Unfortunately this comes at a cost. Those whose income takes them above the

    MIG level lose Pension Credit as it is phased out the higher a pensioner?s

    income. This introduces a new disincentive to save for a far wider range of

    pensioners. And the very complexity of the new system is so bewildering that it

    is bound to undermine the savings culture. The summary of the state pension

    system in Annex A will give the reader a taste of this complexity.

    The Pension Credit works roughly as follows. It will have two components.

    The first is the Guarantee Credit which is simply the new name for the benefit

    needed to bring a pensioner?s income up to the MIG level. The second component

    is called the Savings Credit. Those pensioners whose private pension is less then

    the gap between the BSP and the MIG will get a Savings Credit of 60p for every

    ?1 of their private pension. They will continue to forego ?1 of Guarantee Credit

    for every ?1 of pension income. So they face a net loss of benefit of ?only? 40p in

    the ?1 for their saving income. In other words they still suffer the same penalty

    on marginal savings as someone paying the top rate of income tax.

    Take a worked example: George has a Basic State Pension of ?77.45 plus a

    private pension of ?10 pw. He therefore foregoes ?10 of Guarantee Credit and

    is entitled to the ?14.65 pw to make his income up to the MIG level of ?102.10.

    But from October 2003 he also gets ?6 of ?Savings Credit?. So his take home

    income will be ?108.10 pw. His ?10 pension will make him only ?6 a week

    better off than someone who has not saved a penny.

    If George had a pension of ?24.65 pw – equal to the difference between the

    BSP and the MIG – he would not be entitled to any Guarantee Credit but could

    claim the maximum Savings Credit. That would be worth ?14.79 which is 60%

    of his additional pension.

    To avoid giving that much Saving Credit to everyone with a pension higher

    than this the Credit is tapered out over higher incomes. For each ?1 of income


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 19

    above that level the amount of Savings Credit to which a pensioner is entitled

    is reduced by 40p.

    Suppose George had a pension of ?34.65 pw – ?10 more than the gap

    between the BSP and the MIG. His entitlement to Savings Credit would then be

    ?4 less than the maximum. So he would receive a Credit income of just ?10.79.

    Entitlement to Savings Credit disappears entirely for those with additional

    pensions in excess of ?61.62 pw. But anyone with a pension up to that level

    finds that every ?1 of pension income that they have saved for has only made

    them 60p better off than if they had saved nothing.

    The disincentive is even worse for those with incomes between ?127 and

    ?140 per week in 2003/04. They face a loss of both Saving Credit and 10 per

    cent tax amounting to a combined withdrawal rate of 46p in the ?2.

    As a result of Savings Credit some 3.8 million pensioners will face a penalty

    on their pension or other income of 40% (plus any tax) in 2003/4. Some 1.3

    million of them will have incomes above the MIG level. The other 2.5 million

    will have pre-benefit incomes below the MIG.3 The number facing this

    disincentive will rise steadily as the MIG is up rated faster than the basic

    pension. The government?s own projections show that by the middle of this

    century two thirds of pensioners will be eligible for Pension Credit.4

    The only way to remove in the longer term the disincentives and

    resentments inherent in this system is to set the new State Second Pension,

    together with the Basic State Pension, at a level at least sufficient to lift people

    above the Minimum Income Guarantee. The need for a system of Saving

    Credits would then disappear. The ABI calculate5 that the basic element of S2P

    will need to be over a third higher than it is currently set if the combined value

    of the Basic and Second State Pensions is to exceed the Minimum Income

    Guarantee. That means that the Lower Earnings Threshold, which is currently

    set at ?11,200 per annum, will need to be set at ?15,000.

    Even this only ensures that people with BSP and the flat rate of S2P at

    retirement will have an income in excess of the MIG. As MIG is up rated in line

    with earnings it could soon exceed their compulsory pension income plunging

    them back into the means tested zone.

    Should pensioners? incomes rise in real terms?

    The government is committed to raising the Guaranteed Minimum Income

    (MIG) in line with average earnings during this parliament. It has indicated its

    intention to do so indefinitely by projecting the MIG forward on that basis.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 20

    By contrast, the government is only obliged to raise the Basic State Pension

    in line with the Retail Price Index (though it has chosen to make discretionary

    increases in excess of that in recent years). The value of each person?s State

    Second Pension will reflect the average level of earnings during their working

    life. So the value of the S2P for each successive year cohort of people reaching

    retirement age will tend to be higher than those retiring the previous year.6

    But once individuals retire their S2P (like their BSP) will only be up rated in

    line with prices. So the Minimum Income Guarantee will steadily overtake the

    combined value of the two compulsory pensions for those already retired even

    if they exceeded the MIG at the date of retirement. This will reintroduce the

    disincentive and resentment effects. The only way to avoid means tested

    benefits overtaking contributory benefits is to up rate both by the same


    The populist arguments ? that pensioners should not be left behind and

    that they have a right to share in rising prosperity – naturally favour up rating

    both means tested and contributory pensions in line with earnings. However,

    since the early 80s governments of both parties have shied away from any such

    commitment on the grounds that:

    ? it would be hugely expensive and would therefore impose a significant

    tax burden on the productive economy,

    ? each generation is entitled to the standard of living that they have

    earned, they do not have a ?right? to the wealth created by the next


    ? if people wished to have incomes rising in real terms during their

    retirement they could tailor their pensions to achieve this but nobody

    does so7, and

    ? even if it were possible to up rate state pay-as-you-go pensions to

    reflect rising earnings it would not be feasible to up rate the supposedly

    equivalent private pensions of those who have opted out of the S2P.

    It is therefore better that both compulsory pensions and state retirement

    benefits should be statutorily indexed to prices and any increases in excess of

    that should be made on a discretionary basis if and when government finances

    permit. In practice the Basic State Pension should be increased by the same

    cash amount as the MIG. Thus if the MIG is raised by ?5 more than inflation


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 21

    then the BSP would also be raised by ?5 and not by the same percentage. That

    would avoid the MIG getting out of line with the BSP plus S2P without

    needing to raise the S2P in real terms. It would be undesirable to make

    discretionary increases in S2P since that would be unfair to those who had

    contracted out into private pensions.

    Should the coverage of S2P be extended?

    At present S2P does not cover those earning below the Lower Earnings Limit

    (LEL) and the self-employed.

    The LEL is set roughly equal to the Basic State Pension (BSP). It would be

    strange to force people earning less than the BSP during their working life to

    save in order to have a higher income than the BSP in retirement.

    The self-employed have been excluded for two reasons. First, they often see

    their business as their provision for retirement and want to use any spare

    money they have to invest in it. In practice very few people fund their business

    out of current savings. Moreover only a minority of the self-employed do end

    up with a saleable business capable of funding their retirement. Second,

    reported incomes of self-employed people are a notoriously inadequate guide

    to their real disposable incomes. If true that merely means that the level of

    contributions they would be required to make would be too low to provide a

    commensurate standard of living in retirement. But an inadequate Second

    Pension would be better than no provision at all. At present only a minority of

    self-employed people do make sufficient provision to avoid reliance on means

    tested benefits.

    So the time has come to require self-employed people to make provision

    for a second pension on the same basis as employees.

    Should the second pension be earnings related?

    It is easy to justify the state requiring people to contribute towards a pension

    sufficient to float them clear of means tested benefits. But why should people

    be compelled to provide for additional pension income proportionate to their

    earnings? The original Beverage system did not relate pensions (or any other

    benefits) to the level of earnings.

    This came in with the Graduated Pension, then the State Earnings Related

    Pension Scheme and persists in the State Second Pension. Presumably the

    rationale is: the higher the standard of living people are used to during their

    working lives, the harder they will find it to make do on the Basic State


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 22

    Pension. Yet most people are short sighted. They do not realise that until it is

    too late. So they must be compelled to contribute towards a pension

    proportionate to their earnings during their working life.

    On the other hand all sorts of people face sharp falls in their standard of

    living during the course of their lives. Yet the state simply ensures a basic

    minimum level of income regardless of their previous income. It does not

    require anyone to insure for an extra cushion related to their previous

    earnings. By definition the better off people are the more they can afford to

    save. Arguably,8 in a free country it is up to the individual to choose how much

    pension they provide beyond the mandatory minimum.

    The Labour government has talked of the State Second Pension becoming

    flat rate once it has settled in. On closer reading, however, they envisage that

    those with earnings greater than the Lower Earnings Threshold (currently

    ?11,200 pa) will continue to be required to provide for an earnings related

    second pension (financed by rebates from their national insurance


    My proposals assume that the Mandatory Funded Second Pension includes

    an earnings related element based on the State Second Pension. However, it is

    not an essential part of the scheme. The key component is the flat rate element

    of the State Second Pension. This must be set at a level which, together with

    the Basic State Pension, is at least sufficient to avoid reliance on means tested

    retirement benefits.

    1 “The Future Regulation of UK Savings & Investment”, September 2001, Oliver Wyman & Co/ABI.

    2 Department of Work and Pensions reply to PQ from Peter Lilley MP 24th February 2003 Hansard.

    3 Figures provided to the author by DWP.

    4 “Pension Credit: long-term projections” DWP January 2001.

    5 “Adequacy, affordability and incentives: a better future for state pensions” ABI March 2003.

    6 This is quite distinct from the fact that during the introductory generation successive year

    cohorts will have accrued additional years of entitlement.

    7 To buy an annuity equal to the Basic State Pension increasing in line with prices would cost a 65

    year old man ?70,000. If it was increased by 1.5% per annum more than prices it would cost

    about ?80,000. Estimates provided to the author by GAD.

    8 Nigel Lawson did argue this when SERPS was reviewed. See “The view from No 11” by Nigel



    SMF Save Our Pensions Text 16/9/03 12:06 am Page 23

    2. FUNDING:


    The previous chapter discussed the minimum pension provision people should

    be required to make during their working lives. An equally important question

    is how much of that provision should be funded by saving and investment and

    how much should be left to be paid by future taxpayers?

    The government has recently reaffirmed its commitment that by the

    middle of this century 60% of all pension liabilities should be funded? against

    40% at present. They stand no hope of reaching that target without decisive

    action to shift people from reliance on the state pay-as-you-go pension

    system to private fully funded pensions.

    When the state provides pensions it invariably funds them on a pay-asyou-

    go basis. This year?s taxes are used to pay for this year?s pensions. Nothing

    is saved or invested for the future. That is the method adopted by most of our

    continental partners to pay for the bulk of their pensions. So as their

    populations live longer they face the nightmare of higher taxes falling on a

    declining number of people of working age to pay for their pensions.

    By contrast, in this country we have encouraged people to save and invest

    for pensions beyond the basic minimum. We have enabled employees to

    contract out of the Additional State Pension into occupational or personal

    pension schemes. A majority of eligible employees have done so. When

    employees contract out, they receive rebates from their National Insurance

    Contributions which are saved and invested in industry to earn the profits that

    will eventually pay their pensions – without any further burden on taxpayers.

    Meanwhile this generates an enormous flow of committed long-term

    investment. As a result the total amount invested to meet Britain?s future

    pension liabilities is greater – not just than that invested by any other EU

    country – but than all the rest of the EU put together.

    Nonetheless those funds still only represent 40% of Britain?s future pension

    liabilities. The other 60% will fall on future taxpayers. To meet the

    government?s target of reversing those proportions by the middle of the

    century will require a substantial shift in the number of people contracting out

    of the Additional State Pension system.

    At present employees are given the choice of whether to remain in the


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 24

    unfunded Additional State Pension or to contract out into a funded private

    pension scheme. Those who remain in the state system accrue rights each year

    to an additional pension when they reach the State Pension Age. Those who

    contract out receive a rebate from their National Insurance Contributions

    payable directly into their personal or occupational pension fund. The size of

    the rebate is set on the basis of the Government Actuary?s estimates of the

    amount needed to fund a private pension broadly sufficient to replace the

    pension entitlement that employees would have accrued in the State system.

    The most straightforward way to shift the proportion of employees with

    funded second pensions would be simply to withdraw, at least for new entrants

    to the labour market, the option of an unfunded State Second Pension. All new

    workers would then be required to join a Funded Second Pension scheme of

    equivalent value to the present S2P, financed by rebates from their National

    Insurance Contributions.

    Encouraging more people to opt for a funded second pension by increasing

    the value of rebates relative to accrued rights would have two serious

    disadvantages. First, it would increase the cost to the exchequer. Second, it

    would put the government in the dubious position of offering what would be

    an inferior alternative bearing its imprimatur. If anyone else did this they

    would be guilty of misselling.

    Given that the government wants more people to have funded private

    pensions it is odd to offer an alternative ? an unfunded State Second Pension

    ? which it no longer wants them to join.

    The reason both this government and its predecessor have nonetheless

    continued to offer employees the option of an unfunded Additional State

    Pension is that contracting out into a personal pension has never been

    attractive for those on low and fluctuating earnings. This is because the lower

    an employee?s earnings the smaller the rebate they receive. But the costs of

    running a fund do not decline proportionately since there are fixed costs in

    setting up and running a fund. So charges tend to absorb a disproportionate

    share of the small rebates received by those on low earnings. To avoid this

    unfairness to those on low earnings employees have been given the option of

    remaining within the unfunded Additional State Pension scheme.

    The government sought to overcome the problem of costs absorbing too

    great a proportion of the rebates of low earners by introducing Stakeholder

    Pensions. Stakeholder charges are limited to a maximum of 1% p.a. of funds

    invested. So charges on small funds can no longer be increased beyond that


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 25

    1% cap to reflect the higher proportionate costs of handling them.

    Unfortunately that means that there is little or no incentive for Stakeholder

    providers to encourage low earners, who would only have small rebates to

    invest, to contract out of the unfunded State scheme. Consequently the

    introduction of Stakeholder pensions has done little to encourage more people

    to contract out of SERPS/S2P. Instead there is a worrying trend for individuals

    to opt back into the State Second Pension. Moreover, companies are

    contracting their employees back in wholesale when they switch from direct

    benefit to money purchase schemes. This reflects the widespread perception

    that the value of the rebate is no longer adequate to fund a pension as

    attractive as the unfunded State Second Pension.

    The problem of low earners? rebates being below the threshold necessary to

    justify the fixed costs of setting up a personal pension scheme was inherent in

    SERPS. As its name indicates the amount of pension accrued was directly

    proportionate to earnings and so therefore was the rebate for contracting out.

    The Conservative government consulted on withdrawing the option of

    remaining in the unfunded SERPS. But it concluded that there was no way

    round the problem of earnings related rebates and fixed costs. So it decided to

    retain the unfunded SERPS option for the benefit of those on low and

    fluctuating earnings. Most of those who have not opted out are indeed on low

    or fluctuating earnings.

    As Secretary of State I reconsidered this issue. I concluded that SERPS could

    not be replaced by a Mandatory Funded Pension in a way that was fair to those

    on low earnings unless the unfunded Basic State Pension was also being

    replaced by a funded equivalent at the same time. Hence the radical plan

    called Basic Pension Plus ? (see Annex B). This was based on the recognition

    that everyone paying National Insurance Contributions is accruing the same

    flat rate amount of Basic State Pension regardless of their earnings (unlike

    SERPS). To enable people to build up a funded pension equal to their Basic

    State Pension everyone would be entitled to the same flat rate rebate. This

    would need to be quite substantial ? roughly ?12 p.w. at today?s levels ? well

    above the threshold needed to justify the fixed element of the costs of setting

    up and running a personal pension fund. The same fund could also receive and

    invest, with no need for additional charges, employee?s earnings related

    rebates – however small – to provide a funded equivalent of SERPS. So Basic

    Pension Plus involved giving everyone, as they reached working age, their own

    fund into which would be paid rebates from their National Insurance


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 26

    contributions sufficient to provide a pension at least equal to their Basic State

    Pension plus their State Earnings Related Pension.

    This proposal was traduced during the election campaign which made it

    difficult for the incoming government to develop the idea despite adopting

    some of its rhetoric. However, when the Labour government replaced SERPS

    with the State Second Pension it changed the structure in a way that means

    the problem of low earners receiving rebates inadequate to justify setting up a

    personal pension is no longer inherent to the system. (See Annex A for a fuller

    explanation of the structure of S2P.) It is true that at present the rebates for

    employees contracting out of the new State Second Pension are proportional

    to earnings just as in SERPS. But this feature seems to have been carried over

    from SERPS even though the amount of pension accrued under S2P is no

    longer proportional to earnings for those on low pay. Under the S2P the

    amount of pension rights accrued does not decline if earnings fall below a

    certain threshold (the Lower Earnings Threshold which is set at roughly half

    national average earnings). Employees earning say ?5,000 or ?7,000 will accrue

    the same amount of S2P as someone earning ?11,200 which is the current

    Lower Earnings Threshold. Yet if those with earnings below the LET contract

    out of S2P into a personal pension they still receive a rebate proportionate to

    their earnings. (They continue to accrue an element of S2P based on the

    difference between their actual earnings and the Lower Earnings Threshold.)

    The earnings related rebates mean that it remains unattractive for those on

    low earnings to contract out of S2P.

    The Mandatory Funded Second Pension

    The government could equally well have allowed employees earning below the

    LET to contract out of all the pension rights they accrue under S2P and receive a

    corresponding rebate. Since they accrue a flat rate amount of pension

    entitlement within S2P they should be entitled to a flat rate rebate in respect of

    earnings up to the LET. Such a rebate would be above the threshold level necessary

    to make it worthwhile to set up and handle a personal pension. The problem of

    minimal rebates being absorbed by the disproportionate costs of running small

    funds would therefore disappear. So it would no longer be necessary to provide

    the option of the unfunded S2P for those on low earnings. That option could then

    be withdrawn at least from new entrants to the labour market.

    All new entrants would instead be required to join a Mandatory Funded

    Pension scheme. Those earning less than the LET would receive a flat rate


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 27

    rebate. Those earning more than the LET would receive the same rebate as

    employees who contract out get at present – comprising that flat rate element

    plus an element related to their earnings above the LET.

    The MFSP compared with the government?s plan.

    The government does not appear to have considered this option. It has,

    however, considered ways of restricting employee?s rights to choose between

    the unfunded state system and a funded alternative in Phase Two of S2P. This

    will come in once S2P has bedded down. The government did originally

    consider removing the right to remain in the unfunded S2P from all those

    earning above the LET and requiring them to join a funded private scheme.9

    Conversely, no-one earning below the LET would be allowed to contract out

    into a funded personal pension. They would stay within the unfunded state

    scheme accruing their flat rate entitlement to S2P.10

    Those proposals seem to have survived a last minute rewrite of the White

    Paper. The rest of the White Paper left an element of choice at least for those

    earning above the Lower Earnings Threshold. They will be able to stay in the

    unfunded state system. But if they do so they will only accrue the flat rate

    component of S2P. By contrast those who opt for a funded private pension

    scheme will receive rebates equivalent both to the flat rate and earnings

    related elements of the S2P as at present. They will no longer be able to

    contract back into the state pay-as-you-go scheme.

    Under these proposals there would clearly be a strong financial inducement

    for those earning significantly more than the LET to opt for a private pension.

    But they would not be required to do so. By allowing employees to stay in the

    unfunded state scheme on such disadvantageous terms the government will be

    guilty of blatant ?misselling?. This two tier system will also create a very odd

    situation for those whose earnings fluctuate above and below the LET.

    Moreover, it is rather offensive to treat those on low earnings as second class

    citizens. They will not have their own fund into which they could put any

    additional savings.

    Yet someone earning say half the LET accrues exactly the same entitlement

    to S2P as someone earning the LET. The employee earning exactly the LET will

    either be permitted or required to have a pension fund and will receive a

    rebate sufficient to provide a pension equivalent to the flat rate element of

    S2P. The employee earning only half the LET will have to remain in the state

    unfunded system. Low earners are apparently deemed unsuitable to hold a


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 28

    pension fund or unlikely to use it to make additional savings.

    I propose that everyone entering the labour market after a certain date

    should be treated equally. Over time the whole of the State Second Pension ?

    both flat rate and the earnings related elements – should be replaced by an

    equivalent Mandatory Funded Second Pension. Everyone earning enough to

    entitle them to any S2P should receive a rebate equivalent to that entitlement

    payable into their personal or occupational pension fund.

    Benefits of a Fully Funded Second Pension

    The benefits of such a scheme would be six-fold.

    i Meeting the government?s target

    It would enable the government to meet its target of reducing the

    unfunded element of the nation?s pension liabilities to 40%.

    How rapidly it did so would depend on the speed with which it was

    phased in.

    To do so for all employees not currently contracted out would cost

    over twice as much as the rebates of those currently contracted out.

    See Tables 1 and 2. These figures reflect the current levels at which S2P

    accrues. If this flat rate element was raised as suggested by the ABI to

    lift everyone above the level of the Minimum Income Guarantee

    indexed to earnings, the rebates for the flat rate element would

    increase correspondingly. This would be partly offset by a reduction in

    the earnings related element unless the upper earnings limit was also


    It might be more feasible to introduce the scheme by age cohorts.

    The most gradual way would be (as proposed for Basic Pension Plus) to

    require all young people when they get their first job to have a personal

    fund into which their rebates would be automatically paid. That would

    involve additional rebates mounting initially at a rate of less than ?200

    million11 p.a. cumulatively.

    In practice it should be possible to include within the scheme from

    the start everyone below the age of, say, 30. The initial value of rebates

    to fund pensions at the existing level of S2P would then be ?2.2 billion

    more than at present, mounting by an additional ?35013 million p.a.

    That compares with the current cost of rebates of ?11.5 billion for

    everyone who voluntarily contracts out at present


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 29

    Table 1:

    Value of NI Rebates for those Voluntarily Contracted Out of S2P in 2002/03

    (GB accruals – ?billion)

    Age < 30 Age > 30 All Ages

    People earning below LET12 0.1 0.5 0.6

    People earning above LET 1.3 9.6 10.9

    All earners 1.4 10.1 11.5

    Source: DWP/GAD replies to author

    Table 2:

    Value of NI Rebates for All Earners whether or not currently contracted out

    (2002/03, GB accruals, uncapped APP rebates ? ?billion)

    Age < 30 Age > 30 All Ages

    Those earning below LET

    ? value of rebates for flat rate S2P 1.0 3.6 4.6

    Those earning above LET

    ? value of rebates for flat rate S2P 2.2 13.5 15.7

    ? value of rebates for earnings

    related accruals 0.7 6.4 7.1

    3.9 23.5 27.4

    Source: DWP/GAD replies to author

    ii Encouraging voluntary saving

    As everyone within the scheme would have their own fund they would

    be able to make additional contributions into it as they wish. Nearly

    half of those who contract out of SERPS do save extra on top of the

    National Insurance rebate. The Australian experience suggests that

    since people have been required to have a fund it has become much

    easier to persuade them to make additional voluntary contributions.

    The proportion of those making voluntary contributions on top of the

    compulsory minimum has doubled to 40% compared to 20% previously.

    On average voluntary pension savings now amount to an additional 3

    to 4% of income on top of the 9% compulsory minimum.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 30

    Automatic ownership of a personal pension fund would make it

    much easier for employees to make additional voluntary savings. But

    experience suggests that the proportion who do so is strongly

    influenced by advice and encouragement from their employer.

    Matching contributions from the employer are a powerful incentive to

    save more than the minimum. Research conducted by Watson Wyatt for

    Winterthur14 shows that a programme of education in the workplace

    can have a powerful multiplier effect. Employers who offered to

    contribute an additional 25% of whatever voluntary contributions

    employees make to a 401k plan and backed this with a targeted

    communications programme achieved almost as high a take up as did

    companies who offered a 100% match but carried out no

    communications programme. In the USA full tax benefits on 401k plans

    are only available to company owners or top management if the plans

    are available on the same terms to all employees and a high level of

    take up among all employees is achieved. So managers have an

    incentive to carry out such communications programmes. It would be

    desirable to make similar provision in UK tax law. At present UK

    managers are reluctant to offer any advice to employees even to remain

    in the company pension fund for fear of falling foul of the Financial

    Services Act. This should be revised to give employers the clearest

    possible exemption from such liability.

    iii Reducing costs

    If people automatically have their own fund the costs of running these

    funds will be significantly reduced. The costs of persuasion are largely

    eliminated and a standardised, mass marketed product has lower

    running costs. Moreover, the running costs as a percentage of funds

    under management decline as the amount invested rises over time. The

    Association of Superannuation Funds of Australia estimates that the

    average administration costs would fall from 1.5 % when the average

    balance is US$10,000 to 0.5% when the average balance rises to


    A reduction in the annual cost of running a scheme from 1% of

    capital invested, which is the norm for stakeholder pensions, to 0.5%

    could increase the final pension by some 13%. The Vanguard Group in

    the US charge just 0.3% on their tracker mutual funds.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 31

    iv Spreading wealth

    Every working person would eventually own his or her personal pension

    fund. This would constitute the largest distribution of wealth since the

    spread of home ownership. Working people who earn less than the

    Lower Earnings Limit (currently ?11,200 pa) throughout their working

    lives would end up owning a substantial pension pot. If they put

    nothing apart from their automatic minimum rebate into their personal

    funds they would accumulate a fund worth some ?65,000 in today?s

    prices by the time they reached age 65.16 Someone earning a salary just

    equal to the Lower Earnings Limit who put in additional voluntary

    contributions of 3% of salary ? the average amount contributed

    voluntarily by their Australian counterparts ? would accumulate a fund

    worth over ?115,000. That would finance a second pension indexed for

    inflation equal to nearly two and a half times the Basic State Pension.

    v Freedom of Choice

    Once people have their own fund they have far more freedom to decide

    when they should retire. Yet they will also have the incentive to

    continue working (possibly on a part-time basis) since this will both

    enable them to continue building up their fund and increase the value

    of the annuity they can buy for the remaining years of their life. Studies

    show that those saving for pensions on a defined contribution basis

    work longer than those in defined benefit schemes or solely dependent

    on state unfunded pensions.17

    vi Changing attitudes

    Finally, widespread capital ownership would give people a greater sense

    of having a stake in the economy. Everyone would be more conscious of

    having a vested interest in policies that encourage profitability and

    growth. The sense of ?them and us?, of capital versus labour, and the

    populist idea that we can profit by transferring burdens to some

    abstraction called “business”, would all diminish.

    Arguments against funding all State Second Pensions

    Of course there also are arguments deployed against funding pensions.

    “Funding is pointless” The IPPR18, in particular, argues that there is no

    point in encouraging people to save and invest for their future. They say that


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 32

    people cannot squirrel away a proportion of today?s goods and services like

    nuts. All they do is pile up paper claims on future flows of goods and services.

    In reality tomorrow?s pensions will have to be paid out of the GDP then being

    generated by those then at work. So, they conclude, it makes no difference

    whether tomorrow?s pensioners extract their share of tomorrow?s GDP by

    exercising paper claims held in the books of pension funds or by the state

    imposing taxes reflecting rights accruing in a pay-as-you-go state scheme.

    The key element of this argument ? that tomorrow?s pensions will be paid

    out of tomorrow?s economic output ? is of course true. But if the rest of the

    argument were valid it would follow that the level of saving in an economy is

    irrelevant to the future level of income of its inhabitants. There would be no

    point in saving at all. In addition to arguing openly against the need to

    encourage more saving it is proposed to divert the ?11 billion of NI rebates

    currently saved in pension schemes and spend them on raising current


    The fallacy of this argument lies in the supposition that because saving is

    done via financial instruments it has no effect on future resources available to

    pay pensions. Of course it does. The total level of saving in an economy must,

    by definition, equal domestic investment plus the acquisition of assets

    overseas. Prices and interest rates adjust to bring that about. So extra saving

    must result in more domestic investment or more overseas assets or both. Of

    course, this assumes that extra private saving is not offset, or financed by, an

    increase in public sector borrowing.

    Even if the level of domestic investment were entirely unresponsive to

    more saving, extra savings would be invested abroad. That would mean that

    tomorrow?s pensions could be financed by claims on foreign economies rather

    than taxing our own.

    Finally, levying heavy taxes on the future British economy to pay for

    unfunded pensions may depress activity and drive it abroad whereas

    drawing dividends on investments across the world will not have such an


    “Funded investments are too risky”. Some argue that recent stock market

    falls show that we should not encourage people to invest in the capital

    markets for their retirement. Saving for retirement is a long-term process. Even

    after the recent bear market pension funds show a good return over the long

    term. In any case, it is always a mistake to base policy on short-term

    movements in markets ? still less on extrapolating recent trends. In fact the


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 33

    best time to start broadening reliance on investment in pension funds is after

    a stock market fall rather than at the top of a bull market. The Australians

    believe their compulsory saving scheme got off to a good start precisely

    because it started after the sharp 1987 stock market fall.

    Nonetheless, the greater uncertainty pertaining to funded pensions may

    affect people at different income levels differently. Low earners might be

    expected to be more risk averse than the better off. That is why I believe the

    state should guarantee the flat rate element of S2P. If for any reason the

    portion of a person?s fund built up from their flat rate rebates does not reach

    the level necessary to finance a pension equal to the S2P, the state will top up

    that person?s second pension to the level of S2P they would have accrued. (The

    extra earnings related rebate which people get if they earn above the Lower

    Earnings Threshold would not carry a guarantee. So fund managers would

    notionally divide each person?s fund into a flat rate and earnings related


    Two objections to such a guarantee have been put forward. The first is that

    the state could not afford the cost if markets did very badly. In practice, the

    guarantee would only be called in respect of part of the pension for some

    pensioners in a specific age cohort. By definition the cost to the taxpayer of

    supplementing part of the pension of a section of pensioners cannot be greater

    than the cost of paying the entire pension on a pay-as-you-go basis.

    The second objection is more material. It is that, on its own, a guarantee of

    that kind could lead to investment in excessively risky securities. Fund

    managers would know that if the investments failed, their clients? pensions

    would be topped up by the state. Yet if the high risks produced high rewards

    they would get more than the standard pension.

    I propose that fund managers would be required to invest on a Prudent

    Person basis, as if that guarantee did not exist. Such rules already exist. They

    can be enshrined in statute as in Australia which has largely prevented such

    imprudent investments. The guarantee would only apply to the part of each

    person?s fund built up from their flat rate rebates. Fund managers would also

    be investing the earnings related rebates which would not be guaranteed.

    Managers could be required to adopt the same risk profile for both parts of the


    Means tested benefits already in theory provide a similar sort of guarantee

    and incentive to invest in high risk investments for people approaching

    retirement with a pension fund insufficient to lift them clear of the means


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 34

    testing trap. Yet there is no evidence that financial advisers or fund managers

    do put such clients? investments in high risk assets in practice. So this problem

    may be more theoretical than real.

    “Pay-as-you-go reflects the contract between the generations”.

    Continental politicians invoke this argument to justify their reliance on future

    taxpayers to pay for the pensions they have promised. They say, in effect, that

    parents bear the cost of raising their children and in return have always looked

    to their children to support them in old age. In mature societies each

    generation likewise looks to the next to support it in retirement. That

    argument no longer works. If such a contract ever existed this generation has

    broken it. They are having far fewer children and are living far longer than

    their parents. They have little moral claim on a diminished future work force to

    support them in retirement for far longer than they supported their parents?


    “One generation will pay twice”. Some argue that, although it is desirable

    that everyone should save for their retirement, we cannot reach that goal

    without making the transitional generation pay twice. The current generation

    will have to pay for their parents? pay-as-you-go state pensions while also

    building up savings for their own funded pension. That was an issue I had to

    deal with in my proposal to move to a fully funded basic state pension (see

    Annex B). But it does not apply in the case of the State Second Pension. If

    anything, the reverse is the case. The S2P is a new benefit and is partly funded

    through voluntary contracting out (as was SERPS, which was still building up).

    If previous generations wanted to have a pension in addition to their Basic

    State Pension they had to save for it. They did not pass the obligation on to the

    next generation. That is what we are doing to the extent that we give people

    rights to an Additional State Pension without requiring them to make genuine

    savings and investments to pay for it.

    “Means testing makes it pointless for the poor to save”. The Institute for

    Fiscal Studies has highlighted this problem. But their point is that the current

    level of means tested benefits for retired people makes it uneconomic to save

    voluntarily for a small pension on top of the compulsory state pensions. I

    propose a change in the balance between compulsory minimum provision and

    means tested benefits that will eliminate that problem. Under the present

    arrangements people are already compelled to set aside a certain amount for a

    second pension. As far as this compulsory provision is concerned the

    disincentives are irrelevant. Nor do the disincentives affect whether the


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 35

    compulsory provision for a second pension should be via a state pay-as-you-go

    scheme or a private funded one.

    9 “the proposed system?compulsory funded pensions for those earning over [the LET]” page 105:

    A new contract for welfare: Partnership in Pensions Cmd 4179 December 1998.

    10 ibid “..we expect the new State Second Pension to become a flat rate scheme for those on lower

    earnings, with those on moderate and higher earnings joining a funded pension (with

    contracted out rebates continuing to be earnings related).”

    11 Estimated from figures provided by DWP/GAD for total costs of rebates for year cohorts of

    different ages whether or not currently contracted out.

    12 For people earning below the LET and for people contracted out through occupational schemes

    who earn below the HET (?24,000pa in 2002/03), the rebate does not reflect the full value of

    S2P. S2P top ups are paid in retirement in order to compensate for this. Expenditure on such top

    ups is excluded from the figures shown.

    13 See footnote 12.

    14 Response by Winterthur Life UK Ltd to DWP Green Paper “Simplicity, Security and Choice”.

    15 Schieber SJ & Shoven JB: “Administering a Cost Effective National Program of Personal Security

    Accounts” NBER, Cambridge MA, December 4, 1998.

    16 Assumes earnings and thresholds rise 1.5% p.a. faster than prices.

    17 Prof Blake, The Pensions Institute.

    18 A New Contract for Retirement, R.Brooks, S. Regan & P. Robinson, IPPR.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 36



    If the government chooses not to replace the pay-as-you-go State Second

    Pension by a Mandatory Funded Second Pension, the only way they can hope

    to meet their 60% funding target is by raising the State Pension Age.

    They have already begun to debate this issue by pointing out the truism

    that the only solution to the pensions crisis is “either to save more, to work

    longer or a mix of both”19 The government is also trying ?to lodge the idea in

    the public consciousness that retirement at 70 will mean a higher pension?, as

    Alan Pickering has pointed out.

    If the government goes down this route, the sensible approach would be to

    index the State Pension Age to life expectancy. On average, life expectancy of

    those aged 65 has increased by about 1 month a year over the last twenty

    years. If the government raises the State Pension Age at that rate in future,

    someone now aged 21 would have to work until they were aged 68 years and

    8 months before drawing their state pension.

    Ideally, however, the government should not be telling people how long

    they must work ? still less that they must work ever longer. The great

    advantage of enabling everyone to build up their own pension fund is not just

    that it helps meet the government?s funding target without raising the SPA. It

    also leaves people free to decide when to retire once their fund is adequate to

    provide a decent pension (i.e. greater than the Minimum Income Guarantee).20

    Moreover, it gives everyone a double incentive to work and save longer. Every

    extra year they work gives them a year longer to save and a year less

    retirement to finance, enabling them to buy a larger pension for a given sum.

    Those advantages of working longer are substantial. Take a fairly typical

    man whose pay rises at 3% p.a. in real terms until he is 50 and then remains

    flat. If he wants to retire at 65 on an indexed second pension worth half his

    final salary he would need to save 12.1% of his salary each year throughout his

    working life. However, if he intends to keep working until age 70, he will need

    to save only 7.5% of his salary. Put another way, the cost of an indexed single

    life annuity worth ?10,000 p.a. at age 65 is ?195,000. Buying the same annuity

    at age 70 would cost ?156,000.21

    One way the government could reinforce that incentive to delay


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 37

    retirement, without altering the State Pension Age, would be to use the tax

    system. The major tax advantage of saving for a pension is what Nigel Lawson

    called “the anomalous but much loved lump sum”. At present up to 25% of a

    personal pension fund (excluding that portion derived from NIC rebates) can

    be withdrawn tax free as a lump sum. The lump sum can be spent as the saver

    wishes. It does not have to be used to buy a retirement income. The age at

    which that valuable privilege may be exercised could be raised in line with

    average longevity. People would still be free to withdraw their funds, including

    the tax free element, before that age as long as they used all the money to

    provide an annuity or draw down income sufficient to avoid becoming

    dependent on means tested retirement benefits.

    A number of criticisms could be made of simply leaving people free to

    decide whether to cease working once their fund is adequate to escape means

    tested benefits:-

    i Most people will opt to retire and take their pension at the earliest

    permissible date even if their pension is meagre. Only 8% of men work

    beyond their State Pension Age of 65 though nearly a third of women

    retire after 60.22 It is certainly true that very few people avail

    themselves of the current option of deferring the state pension even

    though they can thereby increase their future state pension. The

    increase is one seventh of 1% for every week deferred, which is about

    7.4% for each year of deferral, up to a maximum of 5 years. The lack of

    interest is partly because this option is little known. Also the

    enhancement is not particularly generous. The current terms are

    roughly neutral on an actuarial basis. They will therefore be less than

    neutral for those with below average life expectancy. That is why I

    decided to improve the terms to one fifth of 1% per week (10.4% for

    each year?s deferral) and to remove the limit on the number of years of

    deferral. The government?s decision to bring that forward is welcome.

    The experience of the self-employed suggests that when people

    own their personal pension fund the double incentive to work longer is

    effective. Most self-employed people own a money purchase pension

    fund. In recent years their average age of retirement has increased

    relative to that of members of Defined Benefit Schemes, including

    those relying on SERPS.23


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 38

    Fig 1. Employment, unemployment and inactivity rates for men aged 50-64;

    UK; spring 1984 to spring 2001

    ii People in manual/stressful/low paid jobs do not want to, and often

    cannot, work past the State Pension Age. To those engaged in

    stimulating jobs the idea of working beyond 65 might be positively

    appealing even if it had no impact on their eventual pension. But is it

    reasonable to expect the man operating a pneumatic drill to go on

    working into his late sixties? Nearly half of all men have ceased to work

    by age 62.24 Early withdrawal from the labour market is particularly

    prevalent among the less skilled. 41% of men aged 50 to 64 with no

    qualifications were inactive compared with 22% of those with GCSElevel

    or higher. Three quarters of men aged between 50 and 64 who are

    economically inactive say they do not want a job. 62% of plant and

    machine operatives gave their reason for inactivity as being long-term

    sick. By contrast less than a quarter of managers and professionals were

    long-term sick and over 55% said they had retired ? often with an early

    occupational pension.25











    2000 1998 1996 1994 1992 1990 1988 1986 1984

    In employment

    Economically inactive

    ILO unemployed

    SMF Save Our Pensions Text 16/9/03 12:06 am Page 39

    Even if the manual groups had possessed a personal pension fund,

    the prospect of adding to it would have been unlikely to persuade many

    of them to work longer to build up a better pension. However, the

    reluctance of lower earners to work up to, let alone beyond, the State

    Pension Age is already a problem under the present system. The changes

    I have proposed so far would not solve that problem but nor would they

    make it worse. The solution is to improve the working of the labour

    market so that people in their 50s can find jobs which are neither

    physically stressful nor demanding of lengthy experience.

    There are signs that this is happening both in the USA and the UK.

    In the USA, the decline in participation rates among men aged

    between 60 and 64 halted in the early 1990s and has begun to reverse

    since then particularly among ?high school drop outs? ? the group with

    the lowest participation rate26.

    Likewise in the UK the employment rate among men aged 50 to 64

    has risen by 5 percentage points since its nadir in 1993.27

    Given appropriate job opportunities, the possession of a personal

    pension fund may well encourage people to defer retirement or to take

    on part-time jobs to supplement an early retirement pension.

    iii Lower income/manual workers have lower than average life expectancy

    so they subsidise the pensions of longer-lived, higher income groups.

    This is true of the pay-as-you-go state scheme as well as private funded

    pensions. At age 65 the life expectancy of an unskilled male manual

    worker is over four years less than that of a male professional.28 If both

    types of employee pay premia into the same fund over their working

    lives, the typical professional will draw out four years more pension

    than the manual worker.

    That is not merely inequitable, it is positively perverse. The less well

    off are subsidising the rich.

    The situation is accentuated at present by the fact that the

    annuities market is dominated by the better off who happen to be


    Annuity rates reflect this since they are calculated on the basis of

    the mortality of actual annuitants rather than the average population.

    As a result an average member of the population would have to pay 7

    or 8% more for an annuity than their average life expectancy warrants.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 40

    Fig 2. Life expectancy for manual and non-manual groups

    Making personal pensions universal and mandatory would

    automatically help remedy this. It would do so by bringing the

    characteristics of those buying annuities into line with the total

    population. However, the fundamental problem of the less well off

    subsidising the longer-lived better off would persist.

    Funded pensions would also make it possible to do something about

    that. In the first place annuity providers could be encouraged to relate

    annuity rates to the life expectancy of different income or social

    groups.29 At present “insurers in the UK rarely use socio-economic

    factors in pricing” annuities.30

    Those who have lowest life expectancy ? generally those on lowest

    incomes ? should get the best annuity rate when converting their fund.

    Annuity providers do offer advantageous annuity rates to smokers and

    higher ?impaired life? annuities to individuals with specific life










    1972-76 1997-99 1972-76 1997-99







    at age 65



    17.4 17.4


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 41

    threatening conditions. Some providers also vary their rates according

    to the size of fund converted into an annuity ? because wealthy people

    with larger funds tend to live longer. Anecdotal evidence suggests that

    the cost per ?10,000 in the fund of buying a large annuity (say ?50,000

    p.a.) can be up to 19% more than for a small annuity (?2,000 p.a.).

    Some 10% of the total annuity market (20% of the open market

    business) is already written on special bases like these. This element is

    rising rapidly.

    However, annuity providers do not yet offer differential annuity

    rates related to past earnings or socio-economic group. It is not clear

    why they do not. It may partly be that the providers do not have

    information on individuals? life-time earnings. To obtain and verify it

    would involve disproportionate cost. The government might be able to

    overcome this by making available information on historic earnings

    from the National Insurance Recording System. This could possibly be in

    a form equivalent to a tax code.

    Another option would be to encourage the development of pension

    funds limited to members of particular occupations. Some funds

    limited to people employed in a specific industry already exist –

    notably the Buildings and Civil Engineering Benefit Schemes (B&CE),

    and also the Heating and Ventilation Engineers. The B&CE is the largest

    supplier of stakeholder pensions. It guarantees zero charges on these

    policies up to 2006. The Australian system is largely based on industrywide

    funds. These pool the contributions of all employment levels

    within each industry. Pooling of investment funds on an occupational

    basis will only be advantageous for members if the average life

    expectancy is below that of other insured lives and if the scheme

    provides annuities which reflect that. B&CE for example can offer its

    members annuities some 18% higher for a given sum than is the norm

    in the open market.

    Schemes limited to manual workers could result in them receiving

    an annuity perhaps 30% higher per ?1,000 of savings than a manager.

    It is not at all clear why such annuities are not on offer. This may be

    because the providers have felt that such provision would be politically

    unacceptable. If so they need to be persuaded otherwise.

    A third option is for the State to relate employees? rebates to the life

    expectancy of their income group. That would mean paying larger


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 42

    rebates to those on low incomes. Under the State Second Pension

    accrual rates are biased in favour of those with earnings below the

    lower earnings threshold (currently ?11,200). However, this is not fully

    reflected in the rebates available for the few who contract out of the

    state scheme in that income bracket.

    19 Hansard col 403 13 Jan 2003 Andrew Smith Secretary of State for Work and Pensions.

    20 Anyone wishing to use their fund to retire before the SPA would be required to buy a fixed term

    annuity for the period up to the SPA equal to at least the MIG followed by an annuity equal to

    the difference between the MIG and the BSP. Legal & General have advocated permitting

    purchase of such fixed term annuities: L & G?s response to “Modernising Annuities” 9th April


    21 The calculation assumes net real return on savings of 4% per annum; earnings rise by 3% in real

    terms between age 21 and 50 then remain stable in real terms until retirement at 65, or 70 if the

    individual continues working. Typical annuity rates are taken from the FSA comparative tables

    for a male non-smoker single life: ?100,000 buys ?5,124 p.a. at 65 and ?6,429 at 70.

    22 “The Dynamics of Retirement: Analyses of the Retirement Surveys” by Richard Disney, Emily

    Grundy and Paul Johnson. DWP Research Report No. 72.

    23 Also see “The Impact of Wealth on Consumption and Retirement Behaviour in the UK”, David

    Blake, The Pensions Institute.

    24 Hansard 13th May 203 DWP written answer to Peter Lilley MP.

    25 “Patterns of Economic Inactivity Among Older Men” Catherine Barham, ONS Labour Market

    Trends June 2002.

    26 ?The Puzzle of Later Male Retirement? Richard Johnson Economic Review, Q3 2002, Federal

    Reserve Bank of Kansas City.

    27 “Patterns of Economic Activity among Older Men” Catherine Barham, Office of National

    Statistics in Labour Market Trends June 2002.

    28 The gap between Social Classes I and V widened from 2.6 years to 4.1 years between 1972-6 and

    1997-9. “Trends in Life Expectancy by Social Class” ONS 18th February 2002.

    29 Annuities purchased with the compulsory element of pension savings must be priced on a

    unisex basis. I do not propose any alteration in this.

    30 ibid page 5.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 43


    The recent gloom about the cost of providing pensions has focussed on the fall

    in the stock market and the increase in longevity. There is little the government

    can do about the former or would want to do about the latter. Remarkably

    little thought has been given to reducing the non investment costs of

    providing pensions. Yet this is the one area where improvements could be

    made which could ameliorate the position.

    Pensions involve two stages: the accumulation of a pension fund and the

    subsequent payment of an annuity or pension. The first stage has by far the

    highest costs. But there are also opportunities to improve value in annuity

    provision to which very little attention has been paid.

    Costs of accumulating pension funds

    The costs of accumulating pensions can be surprisingly high ? especially on

    personal pensions. A study in 199931 put the total cost at some 36% of the

    accumulated balance over a working lifetime. A significant proportion of this

    cost came from the cost of transferring accounts, opening duplicate accounts

    and ceasing to contribute to paid up accounts. Even the 1% per annum cap on

    costs which providers of stakeholder pensions are permitted to charge can still

    reduce the accumulated balance over a working life by between 20% and 25%.

    Put another way, a charge of 1% per annum on funds in a personal pension

    would absorb nearly a quarter of the real return of 4% per annum projected by

    the Government Actuary.

    The factors which have tended to swell the costs of personal pensions are the

    costs of persuasion, advice, compliance, risk of lapse and/or switching to another

    provider all on top of the actual cost of managing the fund and administering

    each account. An analysis32 of Life Office expenses showed that two thirds of

    their costs are accounted for by the cost of acquiring customers ? especially

    paying commissions. Even mutual funds with commission free sales spend half

    their expenses on acquiring customers. One reason occupational funds have

    lower costs33 is that they do not need to attract members. Australian figures

    suggest that the costs of running occupational schemes are typically half those

    of personal pensions. Preliminary results from a study34 of British occupational

    schemes suggested that their costs, though lower than those of personal

    pensions, are not far below the 1% annual limit set for stockholder pensions.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 44

    The government has recently indicated that, rather than reducing the 1%

    cap on stakeholder fees, it is prepared to consult about raising it. Is there any

    scope for reducing those costs? The last Conservative government initiated a

    number of steps to bring about greater transparency, simplicity and

    competition to bear down on costs. These have been carried forward with the

    Pickering Report, the Sandler Report and the Inland Revenue report on Tax

    Simplification. These make many welcome recommendations but at best they

    will bring about marginal reductions in costs and fees.

    The simple truth is that there is only one way to cut costs ?at a stroke?. That

    is to make it compulsory for everyone to have their own pension fund

    throughout their working lives. Compulsion reduces the need for persuasion ?

    the most costly element in the process; dramatically reduces the need for

    advice; means that the product can be standardised and increases the volume

    of investment over which costs can be spread.

    The Australian experience35 is revealing in this respect. Even though there is

    no cap on costs they are coming down steadily as a proportion of funds under

    management. And they are expected to fall significantly further as the average

    balance rises. The Australian experience is one to learn from – not to copy

    slavishly. They have paid remarkably little attention until recently to the level

    or structure of costs and charges. Their costs could well have come down

    further and faster if they had done so earlier ? for example by automatically

    merging the multiple accounts which arise when people move jobs or by

    imposing a cap on charges.

    Over time it might be possible to reduce the charges on the compulsory

    element of saving in UK Stakeholder type funds from 1% to as low as 0.5% per

    annum. That sounds a small saving. In fact it could boost the value of the

    pension accumulated over a working life by 13%.

    Costs were coming down substantially in the UK even before the cap on

    stakeholder costs was introduced. Table 3 shows that they fell by nearly half

    between 1989 and 1997. It also indicates that very significant scope remains

    for further reductions given that the quarter of companies with lowest costs

    have an expense ratio less than half that of the quartile with the highest costs.

    At the same time, the cap could be retained at a higher level (or

    conceivably removed entirely) on any extra savings above the mandatory level

    that people were encouraged to put into their personal funds. This would

    restore the ?reward for persuasion? which has largely disappeared at the present

    level of stakeholder provision.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 45

    Table 3

    Life office expenses

    1997 1994 1989

    Average expenses ratio* (basis points) 130 192 236

    Expense ratio ? first quartile* (basis points) 62 123 174

    Expense ratio ? third quartile* (basis points) 156 171 270

    Average share of business acquisition in total costs* (percent) 65 69 73

    Average share of commissions in total costs* (percent) 31 32 33

    Sample size 146 157 173

    *weighted average ratio of expenses to funds invested.

    Source: Murthi, Orszag & Orszag March 1999 based on Synthesis Life analysis of statutory returns to DTI.

    Reducing the Cost of Annuities

    There has been great resentment among those reaching retirement in recent

    years about the decline in annuity rates. In 1993 ?10,000 would buy a 65 year

    old male a lifetime income of ?1245 p.a. In 2003 it buys just ?741 p.a.36

    This is partly due to falling interest rates reflecting declining inflation. So it

    means retired people who choose a cash annuity, as most do, will not see the

    purchasing power of their incomes eroded so rapidly.

    The other reason is the sharp increase in assumptions of life expectancy.

    Insurance companies are very good at pooling risks across a population with

    stable characteristics. They are less good at predicting trends in the

    characteristics of the whole population ? still less at coping with possible but

    unpredictable events like a magic bullet cure for cancer.

    All they can do is make provision against trends and events which are

    possible though unlikely. Those provisions have to be factored into the cost of

    annuities. So it is more likely than not that annuitants will have paid for

    greater longevity than they will collectively enjoy.

    The uncertainty about future mortality rates is greatest in relation to the

    more distant future and the later stages of life.

    In pricing an annuity for those currently retiring, the providers can assume

    that the proportion of those aged 65 who will die in the next five years will

    reflect recent mortality rates and trends among 65 to 70 year olds. Any

    significant medical advances affecting the period are likely to be in the

    pipeline and therefore identifiable. But it becomes progressively harder to be

    certain about future mortality rates for older ages. The ages showing greatest

    improvements in mortality seem to be steadily rising.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 46

    The Insurance Regulations 1994 require that the price of annuity liabilities

    “be determined on the basis of prudent rates of mortality”. So annuity

    providers should err on the side of caution ? which increases the cost of

    annuities. Some argue that they may nonetheless be significantly

    underestimating the risk of mortality rates improving.37 If so they risk being

    unable to pay the annuities they have promised, should the proportion of

    people living to a great age increases beyond what they have provided for.

    Either way annuitants will lose out.

    There is therefore a plausible case for government bearing the ?risk? of

    financing pensions for the later years of retirement.

    The situation is analogous to that of financing residential care. When the

    last Conservative government was considering how to encourage greater

    private provision we discovered that the average period spent in residential

    and nursing care was fairly short. But that average included a minority of

    elderly people who spent many years in residential care. The insurers faced an

    unquantifiable risk that this minority and their duration of stay would

    increase. We found that if government accepted the responsibility for funding

    residential and nursing care for those who stayed well beyond the average

    term, insurers could reduce their premia disproportionately and more providers

    would enter the market.

    By analogy, the government could assume responsibility for paying S2P for

    the most elderly ? say, those living beyond 85.38 People would then only need

    to accumulate a fund to buy a 20 year fixed term annuity at their S2P level to

    cover their retirement between 65 and 85. If they wanted a retirement income

    higher than the S2P level they would have to buy an open ended annuity for

    the whole period of their retirement.

    In Chapter 3 I postulated that the government may decide to raise the

    State Pension Age (SPA) by about a month each year in line with the rise in

    average life expectancy. In that case the state should be committed to start

    paying S2P to each cohort 20 years after their SPA, rather than specifically at

    age 85. Everybody would then be required and enabled during their working

    life to build up a Mandatory Second Pension fund to provide for a fixed period

    annuity for the first 20 years of their retirement.

    It is hard to calculate the cost of such fixed period annuities relative to full

    life annuities. The mortality tables suggest that 87% of the cost of an ordinary

    annuity for a 65 year old covers the period to age 85. But the funding need

    should be reduced by more than 13% since providers will no longer need to


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 47

    build up a ?provision? for the major uncertainty relating to the later years.

    It is difficult to tell the actual level of annuity companies? costs and

    provisions against risk. A number of studies have suggested that annuities are

    quite keenly priced. These studies are based on comparing the cost of

    purchasing an actual annuity with the cost of government bonds necessary to

    provide that income stream for a population using published mortality


    This methodology suggests that costs (including provisions) amount to

    about 5 or 6% of the price of an annuity. However, the annuity providers may

    be investing partly in corporate bonds and equities and using the higher yields

    to offset a higher level of costs. If so, the cost margin may be substantially

    higher than 6%.

    So there may be significant scope for cost reductions if government takes

    on the cost of S2P for the over 85s.

    31 ?The Charge Ratio on Individual Accounts: Lessons from the UK Experience? by Murthi, Orszag &

    Orszag (March 1999).

    32 Ibid page 44.

    33 This is well-documented in Australia in ?Superannuation Fees and Competition? by Michael Rice

    and Ian McEwin of Phillips Fox (9 April 2002).

    34 Appendix II of Murthi, Orszag & Orszag (March 1999).

    35 ibid.

    36 NAPF.

    37 “?the new Continuous Mortality Investigation projection basis significantly underestimates

    likely future mortality improvements ? Life Offices writing annuity business on competitive

    terms may be making significant losses ? some pension scheme liabilities may be

    underestimated by as much as 30%.” ?Mortality in the Next Millennium? by Richard Willets FFA.

    38 An alternative way for government to bear this risk would be to issue bonds whose annual

    coupons reflect the proportion of the population of retirement age on the issue date who

    remain alive in each subsequent year ? as proposed by David Blake and William Burrows in

    “Survivor Bonds: Helping to Hedge Mortality Risk” The Journal of Risk and Insurance 2001 vol 68

    no 2.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 48



    Both employees and the self-employed should be required to contribute to a

    Mandatory Funded Second Pension at least sufficient, in conjunction with

    their Basic State Pension, to lift their retirement income above the means

    tested benefit level ? the MIG.

    The Mandatory Funded Second Pension should provide a flat rate element

    at least equal to the difference between the Minimum Income Guarantee

    (MIG) and Basic State Pension (BSP). It could also provide an earnings related

    element of pension mirroring that currently included in S2P.

    It would be virtually impossible for the BSP and Second Pension to keep

    pace with the MIG if government continues to raise the MIG in line with

    average earnings and the BSP only in line with prices. Instead both should be

    statutorily up rated in line with prices and any increases above that should be

    the same additional cash amount for the BSP as for the MIG as and when the

    public finances permit.


    Ultimately everyone in work should have a Mandatory Funded Second Pension

    ? either a personal or occupational pension – into which would be paid NIC

    rebates sufficient to fund a pension/annuity equivalent to the S2P during the

    first twenty years of their retirement. Thereafter the S2P will be paid on a payas-

    you-go basis by the State.

    However, this move to compulsory funding could be phased in. Those born

    less than 30 years before the start of the scheme would be required to have

    such a fund from that point or when they entered employment. Those aged

    over 30 when the scheme starts would retain the current option to remain in

    the state pay-as-you-go second pension or contract out into an equivalent

    funded scheme.

    Cutting Costs

    Requiring everyone to have a pension fund is the most effective way to cut the

    costs of running those funds. It eliminates the cost of persuasion, dramatically

    reduces the need for advice, simplifies and standardises and spreads costs far

    more widely. It should eventually be possible to reduce the cap by up to a half.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 49

    That would increase the value of final pensions by as much as 13%.

    The cost of annuities can also be cut disproportionately if government

    assumes responsibility for paying the Second Pension from age 85. That would

    transfer 13% of the cost of expected longevity to the state. But the cost of

    annuities for a maximum 20-year period are likely to be reduced by

    significantly more than that amount because it would no longer be necessary

    to make provision for unlikely but possible increases in longevity beyond that


    Pension Age

    People would be free to draw an income from their compulsory second pension

    fund earlier than the State Pension Age (SPA) if it is sufficient to provide an

    income above the MIG. They would, however, have a double incentive to go on

    working and saving beyond the SPA.

    Should the government fail to adopt these proposals it will be forced to

    raise the SPA. If it does so the logical approach would be to raise it roughly in

    line with life expectancy, i.e. by one month each year. It could, instead, raise in

    this way the age at which people can draw their tax free lump sum unless they

    are using it to buy an annuity.

    These proposals would make it unnecessary to change the SPA though the

    government might decide to do so anyway. In which case it would be logical to

    increase the age at which the state takes over funding the second pension on

    a similar basis.

    To remove the unfairness whereby those on lower incomes who have

    shorter life expectancy subsidise the better off, longer-lived groups the terms

    of annuities should reflect the relationship between life expectancy and


    To make this possible, the Government should make available to annuity

    providers information from National Insurance records summarising people?s

    lifetime earnings in a form similar to their tax code. Alternatively National

    Insurance rebates could be adjusted to reflect the relationship between

    earnings and life expectancy.

    Funds or annuities which more fairly reflect that relationship will make it

    easier for lower income groups to exercise genuine choice over the age at

    which they retire.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 50

    ANNEX A:


    This summary of how the present system works may at least convey to the

    reader how complex it has become. Much of that complexity arises from recent

    attempts to mitigate the disincentive effects of means testing. Unfortunately

    that very complexity itself undermines the savings culture and causes many

    otherwise prudent people to shy away from making any pension provision at



    The state system for retirement income has three components: means tested

    retirement benefits; the Basic State Pension; and the Additional State Pension

    or rebates to obtain an equivalent private pension.

    Means Tested Retirement Provision (MIG & PC)

    Minimum Income Guarantee (MIG)

    There has been a means tested safety benefit since the welfare state was

    established. Originally it was called National Assistance, then Income Support.

    In 1999 Income Support for those above the State Pension Age (SPA) was renamed

    the Minimum Income Guarantee (MIG) and set at a level higher than

    the Income Support level. In 2003/4 its basic value was ?102.10 per week for a

    single person and ?155.80 for a couple.

    Anyone over the SPA with income below the MIG is entitled to have their

    income topped up to the MIG level (subject to an assets test). Every extra ?1 of

    pension, up to this level, therefore results in ?1 less of means tested benefit.

    Pension Credit (PC)

    To mitigate this disincentive, Pension Credit is being introduced from October

    2003. It will have two components ? the Guarantee Credit which tops people?s

    income up to the MIG level and the Savings Credit to partly compensate those

    whose retirement income (additional to the BSP) results in a loss of

    entitlement to Guarantee Credit. Retired people will be entitled to Savings

    Credit equal to 60p for every ?1 of extra pension in the range between the

    value of the full Basic State Pension and the MIG. For those whose extra


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 51

    pension income takes them somewhat above the MIG level their entitlement to

    Savings Credit is reduced by 40p for every ?1 by which their pension exceeds

    the difference between the BSP and the MIG. So instead of losing ?1 for every

    ?1 of extra pension in the range between the BSP and the MIG people lose 40p

    for every ?1 in a range two and a half times as great.

    Basic State Pension (BSP)

    Everyone in work pays National Insurance Contributions (NICs) towards their

    Basic State Pension (BSP) if their earnings are above a minimum threshold

    called the Lower Earnings Limit (LEL).

    The LEL is by convention set at about the same level as the BSP (since it

    would be odd to force people to save to have a higher income in retirement

    than they have in work). The LEL is ?4,004 in 2003/4.

    Entitlement to BSP builds up on the basis of the number of periods in

    which people make NICs or receive NIC credits (regardless of the amount of

    NICs they pay in any period). It takes 44 years of contributions and credits for

    a man, and 39 years for a woman, to earn full BSP. (The number of years

    required for full pension can be reduced for periods caring for young children

    or disabled relatives.)

    The BSP in 2003/4 is ?77.45 per week for a single pensioner and ?123.80

    per week for a couple. It is up rated each year by at least the rate of inflation.

    Additional State Pension (ASP)

    The first Additional State Pension (ASP) ? additional to the BSP ? was the very

    modest Graduated Retirement Pension introduced in 1961. It was replaced by

    the State Earnings Related Pension Scheme (SERPS) in 1978 which was

    replaced by the State Second Pension (S2P) in 2002.

    State Earnings Related Pension (SERPS)

    The State Earnings Relation Pension (SERPS) covered only employees (not the

    self-employed) who earned above the LEL and were not contracted out.

    Each year employees? earnings between the Lower and Upper Earnings

    Limits were taken into account. That element of each person?s earnings was up

    rated in line with rises in an index of average earnings until they reached the

    State Pension Age.

    The rules for calculating the SERPS pension entitlement have changed over

    time. Originally the SERPS pension was to equal 25% of the average of the best


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 52

    20 years of these re-valued earnings. Subsequently the pension was to equal

    20% of the re-valued relevant earnings averaged over a full working life (44

    years for men, 39 for women).

    Once in payment a SERPS pension is increased in line with prices.

    Contracting Out of SERPS

    Employers running an occupational pension scheme offering benefits

    equivalent or superior to SERPS can contract out of SERPS all their employees

    who are members of the scheme. Likewise any individual employee who takes

    out an Approved Personal Pension (i.e. one offering roughly equivalent or

    greater benefits) can contract out of SERPS. In both cases rebates from

    employer and employee NICs will be paid directly into their pension funds. The

    rebates are calculated by the Government Actuary to be sufficient, when

    invested over a working life, to pay for a pension or annuity equivalent to the

    SERPS rights foregone.

    Rebates payable into Approved Personal Pension Funds were related to the

    employee?s age.

    On average rebates are equivalent to nearly 5% of relevant earnings.

    State Second Pension (S2P)

    Like SERPS, S2P covers only employees (not the self-employed) earning above

    the LEL who are not contracted out. Pension entitlement is also based on each

    year?s earnings between the Lower and Upper Earnings Limit, re-valued in line

    with average earnings up to State Pension Age.

    However, the pension entitlement is not simply proportionate to this

    element of earnings. S2P is very redistributive towards those earning less than

    the Lower Earnings Threshold (LET). This is set at ?11,200 in 2003/4. It is

    intended to be roughly half national average earnings and to rise in line with


    Everyone earning between the LEL and LET is treated as if they earned the

    LET. So they accrue the same amount of S2P. The accrual rate for that element

    is 40% (i.e. twice the previous SERPS accrual rate). The accrual rate for

    earnings above the LET up to the Higher Earnings Threshold (HET) is only 10%.

    The HET is set each year so that the average accrual rate at that point is 20%.

    That is ?25,600 in 2003/4. The accrual rate is 20% for earnings above the

    Higher Earnings Threshold (HET) and up to the Upper Earnings Level (LEL). The

    UEL is the limit beyond which higher earnings do not attract higher


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 53

    contributory benefits or charges. It is set at ?30,940 in 2003/4.

    Entitlement to S2P equals the total of the relevant elements of each year?s

    earnings, re-valued and weighted by accrual rates and averaged over a full

    working life.

    S2P in payment will also be up rated annually in line with prices. It is


    Contracting Out of S2P

    The arrangements are similar to SERPS. However, those earning less than the

    LET can only contract out in respect of their actual earnings. If they do so, they

    still accrue an element of S2P in respect of the notional earnings attributed to

    them ? that is the difference between their actual earnings and the LET.

    Those earning more than the LET can contract out and receive rebates

    equivalent to their total S2P entitlement.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 54

    ANNEX B:


    On 5th March 1997 the Prime Minister, John Major, and I announced a

    revolutionary plan to reform the state pension system called Basic Pension


    The Plan involved three key elements:

    A Personal Fund – all young people entering the labour market would

    be given their own fund.

    Rebates – from their National Insurance Contributions would be

    invested in their fund sufficient to finance their basic state pension (a

    flat rebate of ?9 per week rising with inflation) and a compulsory

    second pension (5% of earnings).

    Guarantee – the state would guarantee that fund holders would receive

    at least their Basic State Pension. If the fund?s performance were

    inadequate for any reason, the state would top up their pension from

    that fund to equal the BSP level.

    We said the plan would bring about the largest extension of personal

    ownership of wealth since the spread of home ownership ? and in so doing

    resolve one of the major issues facing modern governments ? providing

    decent, secure pensions for increasing numbers of elderly people.

    Its aims were to guarantee the Basic State Pension; to enable future

    pensioners to share in economic growth; to give a massive boost to investment

    and ultimately to relieve taxpayers of their biggest burden.

    If the extra investment boosted the average growth rate by just one

    twentieth of one per cent (e.g. from 2.25% to 2.30% pa) it would generate

    sufficient extra tax revenues to be self-financing. In any case the ?double

    funding? cost was to be mitigated by changing the timing of tax relief on

    saving for the generation covered by the scheme from an up front to a PEPs

    basis. Pensions for the generations covered by BPP would therefore be free of



    SMF Save Our Pensions Text 16/9/03 12:06 am Page 55

    ANNEX C:



    Much has been written about funded pension systems in Chile and other Latin

    American countries, Singapore?s centrally operated funded system and new

    systems being developed in former communist countries.

    But these countries have totally different political, institutional and

    economic frameworks from the UK. So any lessons are unlikely to be directly

    applicable to us.

    A far more relevant experience is that of Australia ? a developed country

    that shares our Anglo-Saxon institutions. Moreover, it started from a heavily

    means tested universal system of state pensions rather similar to that which is

    being created in the UK by the present Labour government.

    Australia recognised nearly two decades ago that the disincentives and

    resentment inherent in a means tested state system lead logically to

    compulsory savings.

    Before 1983 Australia had a universal Age Pension funded from taxation and

    set at a low level by OECD standards. It was means tested against both

    income and assets. So it resembled Income Support or the MIG/Pension Credit

    system being developed in the UK.

    It acted as a strong disincentive to save for retirement for those on modest

    earnings. There were generous tax incentives for occupational and personal

    pensions but they were generally sufficient to outweigh the disincentive effect

    of means testing the Age Pension only for those on or above average earnings.

    Even those who did save for a pension were often tempted to draw their

    savings as a lump sum and spend them in order to qualify for the means tested

    state Age Pension and related benefits such as health care cards and

    discounted transport.

    Australia?s current system was phased in over more than a decade under

    three successive Labour governments.

    The first pillar remains the means tested Age Pension funded out of taxation.

    This is set at 25% of Male Total AverageWeekly Earnings. Its value in 2002/03 was


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 56

    A$11,164 per annum for a single person. (This was equivalent to ?78 per week.)

    A single person can have private income from other sources up to A$3,016

    and still receive the full rate of Age Pension. Thereafter the Age Pension is

    reduced by 40 cents for every extra dollar of private income until it phases out

    at A$31,304.

    The second pillar is a system of compulsory contributions paid by employers

    for all employees. Since July 2002 this contribution has been set at 9% of

    earnings. The self-employed are excluded as are most of those earning less

    than a Lower Earnings Limit of A$450 per month. There is also an Upper

    Earnings Limit of about A$90,000 per annum beyond which the contribution is

    not compulsory.

    Some 88% of the workforce are within this system as against 40% with

    private pensions before it began.

    The contributions are paid into each employee?s personal account in a

    privately managed retirement fund. Not-for-profit trustee superannuation

    funds were established on an industry basis.

    The intention is to build up a fund sufficient to buy an annuity worth,

    together with the means tested Age Pension, about two thirds of preretirement

    income. Table 4 shows projections for different contribution periods

    and income levels.

    Table 4 Replacement income generated by Australian compulsory superannuation scheme

    plus means tested Age Pension

    30 years of 40 years of

    contributions contributions

    Total average salary % of gross % of net % of gross % of net

    pre- retirement pre-retirement pre-retirement pre-retirement

    A$20,000 70 79 82 90

    A$40,000 45 55 58 70

    A$60,000 37 48 50 62

    Source: Superannuation over two decades: The politics of pension reform in Australia. Senator The Hon Nick Sherry Jan 2003.

    The third pillar is voluntary savings. It had been anticipated that voluntary

    savings would decline as people assumed that the compulsory level was the

    ?correct? amount. In fact, the level of additional voluntary contributions into

    the superannuation schemes has doubled from 20% to 40%. On the other

    hand employers who had contributed more than the compulsory 9% have

    tended to reduce to that level.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 57


    The cost of administering superannuation schemes in Australia averages 1.2%

    of the total amount invested and is declining as this increases.

    This is comparatively low given that the system is still young, there is no

    cap on costs and there is no central clearing house like the UK National

    Insurance system that collects and distributes the NI rebates in Britain.

    Moreover, the Australian system has unexpectedly generated a multiplicity of

    small accounts as individuals change jobs and industries. On average there are

    2.5 accounts for every member. It is proposed that accounts should

    automatically be consolidated. This should significantly reduce costs further.

    There is no requirement that charges be proportionate to the amount

    invested so those on low incomes are not protected from bearing a

    disproportionate level of costs in some schemes.


    Industry-wide funds are run on a trustee basis with half the trustees appointed

    by the employer and half by the unions/employees. To make a decision requires

    a two thirds majority.

    The ?Prudent Person? rules have been translated from Common Law to

    Statute. They require investment managers to invest at arms length, diversify

    and match risks with liabilities. Company schemes must reduce the amount

    invested in the firm to a maximum of 5% of the total.

    There is a compensation fund providing 100% cover for theft or fraud. So

    far only one fund has been reported involving A$30m out of a total of A$530


    Tax Issues

    As compulsion removed the necessity for tax breaks to encourage saving, the

    tax regime for pension funds has become less generous. Indeed tax is now

    vying with declining administrative costs to become the major cost for pension


    A key and, to UK observers, unusual feature of the Australian system has

    been that savers have no obligation to convert their pension savings into an

    annuity on retirement. They can take 100% as a lump sum and put it to any use

    they choose. There is now a tax charge on exit, but savers can still take the

    whole lump sum and spend it on anything. The amounts accumulated under

    the new compulsory scheme by those reaching retirement are small so far.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 58

    About two thirds are taken as a lump sum ? albeit often to repay mortgages or

    other debts. Australia may need to change these rules. The clear lesson for the

    UK, however, is that the primary call on retirement savings should be to ensure

    retired people are not dependent on means tested benefits.

    Savings Ratio

    After the introduction of compulsory savings in Australia there were at first

    signs that the long-term declining trend in household savings was beginning

    to come to an end or even reverse. But the downtrend resumed towards the

    end of the 1990s. Other Anglo-Saxon countries experienced a similar longterm

    decline, worsening in recent years. This may reflect a growing willingness

    to borrow as buoyant housing and stock markets increased the value of

    people?s assets. In Australia the knowledge that funds accumulated in

    compulsory superannuation accounts can be used to repay debts may even

    encourage people to borrow during their working lives.

    Public Sector Employee Pensions

    Apparently public sector pension schemes moved from being defined benefit

    to defined contribution, similar to those in the private sector, with remarkably

    little fuss.

    39 Fuller details are available on my website ? www.peterlilley.co.uk ? under Speeches and Articles

    Feb 1999.

    40 The average exchange rate during 2002/3 was A$2.754/?.


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 59


    SMF Save Our Pensions Text 16/9/03 12:06 am Page 60