Save our Pensions

- Sunday, 19th January 2003

 


19/09/2003
--------------------------------------------------------------------------------


THE SOCIAL MARKET FOUNDATION



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



companies.



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



represent a corporate opinion of the Foundation.



Chairman



David Lipsey (Lord Lipsey of Tooting Bec)



Members of the Board



Viscount Chandos



Gavyn Davies



David Edmonds



John McFadden



Baroness Noakes



Brian Pomeroy



Director



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



SAVE OUR PENSIONS



Peter Lilley



SOCIAL MARKET FOUNDATION



September 2003



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



ACKNOWLEDGEMENTS



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



information.



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



CONTENTS



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



6 SAVE OUR PENSIONS



GLOSSARY OF TERMS



(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



EXECUTIVE SUMMARY



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



EXECUTIVE SUMMARY 7



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.



8 SAVE OUR PENSIONS



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.



EXECUTIVE SUMMARY 9



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.



10 SAVE OUR PENSIONS



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



EXECUTIVE SUMMARY 11



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



customers.



? 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



12 SAVE OUR PENSIONS



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



EXECUTIVE SUMMARY 13



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



INTRODUCTION



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



14 SAVE OUR PENSIONS



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.



INTRODUCTION 15



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



1. COMPULSION:



HOW MUCH PENSION PROVISION?



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



chapter.



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



pension".



A surprising number of commentators seem unaware that all employees are



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



pension.



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



benefits.



16 SAVE OUR PENSIONS



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



COMPULSION: HOW MUCH PENSION PROVISION? 17



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



destitute.



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



18 SAVE OUR PENSIONS



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



COMPULSION: HOW MUCH PENSION PROVISION? 19



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.



20 SAVE OUR PENSIONS



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



amount.



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



generation,



? 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



COMPULSION: HOW MUCH PENSION PROVISION? 21



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



22 SAVE OUR PENSIONS



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



contributions).



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



Lawson.



COMPULSION: HOW MUCH PENSION PROVISION? 23



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



2. FUNDING:



SAVE NOW OR TAX LATER



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



24 SAVE OUR PENSIONS



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



FUNDING: SAVE NOW OR TAX LATER 25



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



26 SAVE OUR PENSIONS



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



FUNDING: SAVE NOW OR TAX LATER 27



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



28 SAVE OUR PENSIONS



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



raised.



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



FUNDING: SAVE NOW OR TAX LATER 29



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.



30 SAVE OUR PENSIONS



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



US$30,000.15



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.



FUNDING: SAVE NOW OR TAX LATER 31



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



32 SAVE OUR PENSIONS



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



pensions.



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



effect.



"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



FUNDING: SAVE NOW OR TAX LATER 33



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



element.)



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



fund.



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



34 SAVE OUR PENSIONS



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?



generation.



"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



FUNDING: SAVE NOW OR TAX LATER 35



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.



36 SAVE OUR PENSIONS



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



3. PENSION AGE:



HOW LONG TO WORK?



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



PENSION AGE: HOW LONG TO WORK? 37



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



38 SAVE OUR PENSIONS



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



PENSION AGE: HOW LONG TO WORK? 39



0



10



20



30



40



50



60



70



80



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



longer-lived.



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.



40 SAVE OUR PENSIONS



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



PENSION AGE: HOW LONG TO WORK? 41



0



5



10



15



20



12.3



13.1



Men



1972-76 1997-99 1972-76 1997-99



Women



14.6



16.8



Manual



Life



expectancy



at age 65



Non-Manual



16.5



17.4 17.4



19.8



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



42 SAVE OUR PENSIONS



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



2002.



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.



HOW TO CUT PENSION COSTS? 43



SMF Save Our Pensions Text 16/9/03 12:06 am Page 43



4. HOW TO CUT PENSION COSTS?



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.



44 SAVE OUR 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.



HOW TO CUT PENSION COSTS? 45



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.



46 SAVE OUR PENSIONS



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



HOW TO CUT PENSION COSTS? 47



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



projections.



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.



48 SAVE OUR PENSIONS



SMF Save Our Pensions Text 16/9/03 12:06 am Page 48



CONCLUSION



Compulsion



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.



Funding



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.



CONCLUSION 49



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



age.



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



earnings.



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.



50 SAVE OUR PENSIONS



SMF Save Our Pensions Text 16/9/03 12:06 am Page 50



ANNEX A:



HOW THE UK STATE PENSION SYSTEM WORKS



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



all.



STATE PENSION PROVISION



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



ANNEX A: HOW THE UK STATE PENSION SYSTEM WORKS 51



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



52 SAVE OUR PENSIONS



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



them.



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



ANNEX A: HOW THE UK STATE PENSION SYSTEM WORKS 53



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



unfunded.



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.



54 SAVE OUR PENSIONS



SMF Save Our Pensions Text 16/9/03 12:06 am Page 54



ANNEX B:



BASIC PENSION PLUS



On 5th March 1997 the Prime Minister, John Major, and I announced a



revolutionary plan to reform the state pension system called Basic Pension



Plus.



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



tax.



ANNEX C: AUSTRALIA?S EXPERIENCE OF COMPULSORY PERSONAL PENSIONS 55



SMF Save Our Pensions Text 16/9/03 12:06 am Page 55



ANNEX C:



AUSTRALIA?S EXPERIENCE OF COMPULSORY



PERSONAL PENSIONS



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



56 SAVE OUR PENSIONS



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.



ANNEX C: AUSTRALIA?S EXPERIENCE OF COMPULSORY PERSONAL PENSIONS 57



SMF Save Our Pensions Text 16/9/03 12:06 am Page 57



Costs



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.



Security



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



billion.



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



funds.



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.



58 SAVE OUR PENSIONS



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/?.



ANNEX C: AUSTRALIA?S EXPERIENCE OF COMPULSORY PERSONAL PENSIONS 59



SMF Save Our Pensions Text 16/9/03 12:06 am Page 59



60 SAVE OUR PENSIONS



SMF Save Our Pensions Text 16/9/03 12:06 am Page 60


Go Back

 

 

Current Location:

 

Home / Publications / 2003

Save our Pensions

 

Search this Site

 

 

Contact Peter Lilley MP:

 

E-mail

feedback@peterlilley.co.uk