Select Committee on Work and Pensions Minutes of Evidence

Memorandum submitted by the Institute for Public Policy Research (IPPR) (PC 27) (continued)


  3.13  We start by establishing the overall cost of the Government's pension settlement. The modelling by Hawksworth (2002) suggests that the net cost of the Pension Credit will rise to 1.2 per cent of GDP by 2050. Overall, the Government's pension settlement will involve direct expenditure costs of 6 per cent of GDP by 2050, compared with 5 per cent of GDP in 2000-01, even if the BSP remains indexed in line with prices. We have discussed earlier how far these costings are compatible with those set out by the Government (DWP, 2002) and this is discussed further in Hawksworth (2002), with the conclusion that they are broadly comparable. These represent of course only the direct expenditure costs: there will in addition be the costs of the "true" tax expenditures and subsidies. These cannot be modelled but in this analysis they are assumed to remain at around the current level of 1.3 per cent of GDP.

  3.14  The key punchline here is that the objective set out in the Prime Minister's forward to the 1998 Green Paper that public spending on pensions as a proportion of GDP would fall, will now not be realised. This change to one key feature of the Government's pensions settlement opens up a very obvious question: had it been known in 1998 that such an increase in spending was possible would we have chosen a different means of delivering the other objectives for the pensions system?

  3.15  There are other key questions that the modelling helps to clarify. The Pension Credit combined with a BSP indexed in line with prices will draw an increasing proportion of pensioners into means-testing with all the consequences for the incentives faced by people and their perceptions of the equity of the system that were discussed above. Even in 2003 around half of all pensioners are expected, according to Government estimates, to be eligible for means-tested support. By 2050 up to 70 per cent of pensioners will be eligible for the means-tested pension credit (Hawksworth, 2002). Falkingham et al (2002) look at an individual on average earnings of around £20,000 a year and with no employment breaks, saving the minimum in a stakeholder pension implied by the level of the contracted out rebate, with a generous 3.5 per cent annual real rate of return, and buying a non-indexed annuity. They would be eligible for the means-tested savings element of the Pension Credit after only three years of retirement. If they purchased an indexed annuity with their retirement savings pot in 2050, they would be immediately eligible for the means-tested savings element of the Pension Credit. If they lived to beyond 80 they would become entitled to the MIG as well.

  3.16  Of much greater concern is that by 2050 anyone with income only from the BSP and the State Second Pension will automatically fall below the MIG threshold on retirement at 65. As Falkingham et al (2002) point out this raises real questions about the purpose of the State Second Pension. Someone who works all their life and makes contributions to the State Second Pension will not have a non-means tested income in retirement that will lift them above the poverty line. It is not clear that this fulfils the objectives for the State Second Pension that were outlined in the 1998 Green Paper.


  3.17  The option of raising the BSP to the level of the MIG and then indexing it in line with earnings would silence most of the Government's critics on the centre-left in one fell swoop as it is what they have been campaigning for over many years. It would be generally very popular in the country as a whole. The option has been dismissed in the past by Whitehall on two grounds: firstly, affordability and, secondly, that it would deliver windfall gains to better off pensioners who do not strictly "need" a higher income from the BSP.

  3.18  Our modelling allows us to confirm the cost of this option. Without offsetting measures, raising the BSP to the level of the MIG by 2010-11 would add around 0.9 per cent of GDP to direct spending on pensions by this date when compared with the Government's base case. This estimate takes into account the savings by not having to introduce the Pension Credit and by fewer people having to claim the MIG top-up. It slightly overstates the extra costs, however, in that the BSP is taxable and 22 per cent and 40 per cent of the extra cost would be clawed back from better off pensioners. This is one way in which the option can be seen to be both more affordable and that the windfall gains to better off pensioners can be mitigated.

  3.19  By 2050 this option would raise direct spending on pensions by about 2.6 per cent of GDP when compared with the Government's base case. Again this does not take into account the lower net costs that would result from some flow of income tax revenue back to the exchequer. Overall direct expenditure on pensions would total 8.6 per cent of GDP in 2050. This is significantly higher than in the Government's base case of 6 per cent of GDP, but is in no sense completely unaffordable. A glance back at the table in chapter 2 shows that it would still leave public spending on pensions in the UK lower than in all other EU countries based on current projections. Simply indexing all taxes in line with price rises and thereby allowing the phenomenon of "real fiscal drag" to increase tax revenue would increase the tax take as a proportion of GDP by 2.3 per cent in just one decade (IFS, 2001). Nevertheless this significant increase in direct spending on pensions by definition carries with it an opportunity cost. The revenues devoted to increasing the BSP would not be available to tackle child poverty, increase spending on health care, education or public transport, or any other of the many priorities faced by Government.

  3.20  This reform option would score highly against the objective of securing an adequate retirement income for all and the objective of maintaining people's incentive to save for their own retirement. With nearly everyone automatically receiving a full non-means tested minimum pension at the age of retirement, we get round the problem of the inadequate take-up of the MIG and thus deliver a significant boost in income to the very poorest pensioners. The MIG would retreat to being a residual means tested benefit for those without a full entitlement to the BSP, a section of the population that we expect to see decline sharply over time.

  3.21  In terms of incentives, the existence of a non-means tested pension in retirement at the level of the MIG would mean that individuals faced a much clearer incentive to accumulate their own savings through pensions or other vehicles. They would know that they were not going to be faced by an ill-understood system of means testing in retirement that might significantly reduce the return to their own savings. However, individuals facing significant housing costs in retirement might still face the high marginal tax rates implied in the housing and council tax benefit regimes, and this is the one significant incentives problem that the gold plated base case does not surmount. The answer here would lie in a fundamental reform of the housing and council tax benefit regimes.

  3.22  It is highly unlikely that any government would simply raise the BSP to the level of the MIG and not take the opportunity to make other changes to the pensions settlement designed to achieve the other main objective of reform: a simplification of the system. From 2007 it is planned that the State Second Pension will become flat rate to be added to the value of the BSP for those who are contracted in. However, with the BSP raised to a level that would be somewhat higher than the Government's planned level for the combined BSP and State Second Pension, the opportunity would arise of simply phasing out the latter altogether. For those on low incomes the BSP would be generous enough in its own right to provide people with a minimum retirement income.

  3.23  Taking the opportunity to phase out the State Second Pension would not only significantly simplify the system, it would temper the "affordability problem" that Whitehall in particular will be concerned about. We can model the effect of phasing out the State Second Pension in the context of raising the value of the BSP. In itself over the short term this would make little difference. By 2010 the revised gold plated option shaves only 0.1 per cent of GDP from the overall cost of the pension reform package. However, by 2050 when the State Second Pension would have matured, its phasing out would lower direct public spending on pensions by 1 per cent of GDP when compared with the full gold plated option. This would leave the option costing overall 7.6 per cent of GDP in terms of direct public spending compared with the 6 per cent of GDP that is the cost of the Government's base case.

  3.24  However, the analysis does not stop there. If the State Second Pension was phased out (perhaps from 2007) there would by definition be no need to retain the expensive system of offering rebates or lower contribution rates to people contracting out of state second pension provision. These rebates cost 0.9 per cent of GDP in 2000-01. We cannot model the cost of these rebates out into the future, but we can use the working assumption that the cost remains the same as a proportion of GDP. In this case abolition of the rebates means that in 2010-11 the total cost of the gold plated reform package as a proportion of GDP would be 5.5 per cent, almost exactly the same as the Government's base case. The reform is basically revenue neutral over the first decade. By 2050 the total cost of the gold plated option would be 6.7 per cent of GDP including the savings on the rebates compared with the 6 per cent cost of the Government's base case not including the rebates. The gold-plated reform option along with the phasing out of the State Second Pension is barely less affordable in the round.

  3.25  It is important to clarify what is going on here and in particular the distributional consequences of what is being proposed. The initial cost of raising the value of the BSP to the MIG by 2010-11 is paid for by closing state second pension provision. In addition we abolish the rebates paid to those opting out of this provision and effectively everyone moves to the contracted-in rate of NICs. The main distributional impact here is to transfer resources between the current generation of workers who would have benefited from having the rebates paid into their pension pots or paying the lower contracted out rates and the current generation of pensioners who will benefit from the higher BSP. It is a transfer from the current workforce to their parents and/or grandparents. Over the long run it is a transfer across the lifetimes of individuals who give up their entitlement to state second pension provision in return for a more generous BSP.

  3.26  Nevertheless some people—and especially those in Whitehall—are going to worry about the "windfall" gains that will go to many of today's better off pensioners that do not strictly "need" the extra income from the enhanced BSP. These pensioners are unlikely to complain about the extra income, but their sons and daughters may complain about losing their rebates or paying higher contribution rates.


  3.27  To what extent could some of the distributional consequences of the reforms be offset through changes to the current tax treatment of pensions and other forms of saving? The major distributional problem that many people might perceive with the gold plated reform package is that it will give a large windfall gain over the next decade to existing middle and high income pensioners, paid for by the current middle and high income workforce who will lose their rebates. Some of these windfall gains would be automatically taxed away at the pensioner's current marginal rate. However, given the existing quite generous tax breaks for the current generation of pensioners, the question arises of whether this distributional impact might be partially offset by reducing the value of those tax breaks. This would achieve some redistribution of the gains within the pensioner population away from more affluent pensioners who would otherwise stand to gain. It would also raise some revenue for the overall reform package.

  3.28  Before we explore the options it is worth setting out the principles that should underlie any changes to the tax treatment of pensions and other forms of saving.

    —  a key objective for any centre-left government is to make the overall tax system significantly more progressive;

    —  saving long-term in a pension needs to remain the most tax favoured form of saving;

    —  an effort should be made to simplify the system if possible;

    —  it would be a bonus if the overall cost of tax expenditures could be reduced to release resources for raising the basic state pension and/or to fund a more generous settlement in long-term care.

  3.29  One modest way of reducing the windfall gains to better off pensioners would be to freeze the age allowances until they were at the proposed level for the BSP/MIG. This might satisfy most people's judgement of an equitable tax treatment in that pensioners reliant solely on the BSP would not be taxed on it, but pension income above this level would be taxed at the normal rates. With the minimum income set at £100 a week from 2003 and then indexed in line with earnings and the age allowances frozen in nominal terms from 2003, it would take only about four to five years for convergence to be achieved. This would imply a clear u-turn in government policy, which is to index the age-related personal allowances in line with earnings from 2003-04 for the remainder of the current Parliament. The proposed reform would generate some additional revenue over the middle of this decade as the BSP was being increased to the level of the MIG, but this gain in revenue would be modest.

  3.30  The additional married couple's age allowance for the over 65s could be abolished outright at an appropriate point. With the BSP being raised in value significantly and an ever-higher proportion of people establishing entitlement to the full BSP in their own right, any justification for the continuation of this allowance would appear to be very weak.

  3.31  Getting rid of the tax-free lump sum is favoured in some quarters, though this would be politically difficult. It could not be done retrospectively of course and so would only affect those "cashing in" their pensions from the date of the reform. Such a change would probably need to be phased in so as not to create a sense of inequity between age cohorts. It would not therefore raise much revenue in the medium term or offset the immediate distributional effects of the proposed reforms. The maintenance of the tax-free lump sum would be one way of ensuring that the tax advantages of saving in a pension vehicle were seen to remain intact.

  3.32  The abolition of higher rate tax relief on pension contributions is attractive on distributional grounds as is abolition of the upper earnings limit on NICs. However, both would further impact on the current workforce rather than existing better off pensioners and so would not solve the perceived problem of the windfall gains to those current better off pensioners. The issue of higher rate tax relief on pension contributions raises important questions about the structure of the tax system that cannot be resolved here.

  3.33  We have already flagged up other reforms to the tax treatment of pensions and other forms of saving, including the abolition of tax relief for equity based ISAs after 2004 and the ending of the loophole that allows grandparents to contribute to their grandchildren's pensions through the stakeholder vehicle. Alongside the freezing of the age allowances and the ending of the married couples' allowance, we have a package of reforms to the tax treatment of pensions and savings that would satisfy the principles set out above:

    —  the tax system would be made more progressive by abolishing equity ISAs that will from 2004 only be benefiting higher rate taxpayers anyway and by scaling back the age allowances;

    —  the abolition of equity based ISAs and the continuation of the tax free lump sum would mean that pensions remained clearly the most tax favoured form of saving;

    —  the abolition of equity based ISAs and the married couple's allowance would simplify the system;

    —  some resources would be released to finance the increase in the BSP or to fund a more generous settlement in long-term care.

  3.34  Some of the windfall gains to current better off pensioners would be tempered. However, it is an inevitable feature of a reform package that relies on increasing a universal benefit that a significant part of the gain will go the better off. This has to be traded off against the other objectives that are secured by the reform package: people receiving an adequate non-means tested pension in retirement; a clearer set of incentives for private saving; and a significant simplification of the system.


  3.35  Several countries have planned for or are considering raising the official retirement or pensionable age at which the state pension is paid. The US is raising its official retirement age to 67 by 2027 and this is also the age that Sweden has chosen. It is interesting that these two countries that are often seen to represent the polar models of advanced market economies are both implementing changes to the official retirement age similar to the one that is considered here.

  3.36  Such a change needs to be justified on more than just the grounds of improving the affordability of the state pensions system. There is one particularly important argument to make. There is evidence that the official retirement age impacts on labour force participation rates (Disney and Johnson, page 10). We could expect an increase in the official retirement age to drag up labour force participation rates in its wake, thus helping to achieve one of the goals that everyone recognises as desirable. Indeed the modelling by Hawksworth (2002) assumes that raising the official retirement age has this desirable impact on labour force participation, making the point that such an administrative change only has the full desired impact if it changes actual behaviour.

  3.37  There are other principled arguments to make. Hawksworth (2002) models the impact of raising the official retirement age over the period 2020-30. This would be the decade after the official retirement age would have been raised from 60 to 65 for women to align it with the age for men, which will occur between 2010 and 2020. This already announced reform would therefore set a useful precedent for raising the official retirement age further and sensitise the whole population to the issue. For the cohort first fully affected—that is those who would have reached 65 in 2030—average life expectancy is now in the middle to late 80s for both men and women. So retiring at 67 would still leave an average of around 20 years of retirement for most men and women. Many people could expect to live for most of that period in a reasonably robust state of health. With an average of around 20 years of childhood and full-time education and training, people would spend around 40 years not in the workforce and around 45 years in it. It is intuitively plausible to think that you have to work for more years than you can live in "leisure".

  3.38  The distributional impact of raising the official retirement age is easy to perceive. In effect the same cohort is paying for its own more generous state pension by being required to work a little longer—a redistribution across the lifetimes of these individuals. Whether the official retirement age is raised or not, the Government should engage the public in a debate about trends towards early retirement in the context of increased healthy life expectancy.

  3.39  There are other practical steps that should be taken. The Government has already recognised that a key priority of the new Jobcentre-plus service that will integrate the Employment Service and Benefits Agency should be to increase activity rates amongst the over 50s—the employment and pensions agendas are indeed inter-linked. The suggestion of the Performance and Innovation Unit (PIU 2000) that people should be able to take part of their pension while continuing to work part time for their current employer needs to be taken forward. And the Government will need to consider carefully the role of future legislation, which it is required to introduce by 2006 to comply with an EU directive, to prevent age discrimination and to promote age equality (Spencer, 2002, forthcoming).

  3.40  Currently, the gold plated reform package that we have proposed would still result in a significant increase in public spending on pensions over the period 2020-40, when the bulge in the retired population occurs. They will now be entitled to a significantly more generous BSP than would otherwise be the case with the BSP indexed only in line with prices.

  3.41  The modelling by Hawksworth (2002) analyses an increase in the official retirement age to 70 over the decade from 2020-30. We have made a pro-rata adjustment to his costings to look at the impact of raising the official retirement age to 67. This would be sufficient to make the whole gold-plated reform package broadly revenue neutral. In this case overall expenditure on the reform package would be around 6 per cent of GDP in 2050, exactly the same as in the Government's base case. In practice of course other elements of the reform package such as the phasing out of the State Second Pension could be traded-off against a larger increase in the official retirement age than the one suggested here.


  3.42  It should be clear that we think that the reform package outlined here with a much greater role for a more generous universal Basic State Pension satisfies the objectives for the pensions system better than the Government's current settlement. The issue of whether adequacy for pensioners is best delivered through a means-tested or a universal approach is an empirical one. The current Government has in effect balanced the two approaches. The MIG has been used to channel resources to the poorest elderly but at the same time the BSP has been increased in real terms and Winter Fuel Payments and Free TV licences (for the over-75s) have been introduced on a universal basis. The 2001 Pre-Budget Report showed that around two-thirds of the overall gains for the average pensioner family from these policies had resulted from increases in universal benefits with the MIG actually playing the junior role (HM Treasury, 2001, page 89). Government policy has thus been more pragmatic and more universalist than would have been expected at the time of the 1998 Green Paper. Rejecting the gold-plated reform on the grounds that it is too universalist would seem strangely ideological for a government that prides itself on implementing what works.

  3.43  The Government's strategy has, however, been essentially ad-hoc, with annual adjustments made to the pension settlement. The political attractions of this approach to the Chancellor might seem obvious. However, it means that people really cannot see how the system is supposed to evolve over the long-term, which is the necessary background against which they can make their own plans for retirement.

  3.44  However, what is being suggested here as part of the gold plated base case—the indexation of the BSP in line with earnings—will raise concerns in Whitehall. Again, however, this cannot be a matter of principle. The 2001 Pre-Budget Report stated explicitly that "The Government is committed to raising the MIG in line with earnings throughout this Parliament" (HM Treasury, 2001, page 88). It also stated that "the age-related personal allowances will be raised in line with earnings rather than prices from 2003-04 and for the remainder of the Parliament" (HM Treasury, 2001, page 92). Earnings indexation is therefore firmly on the Government's agenda. If in principle you can earnings link the MIG and the age allowances then in principle you can earnings link the BSP. Which of these you want to do is an empirical matter and depends upon your view of the affordability constraint. And of course the effect of the ad-hoc increases in the BSP plus the Winter Fuel Allowances is that the universal element of pension provision has broadly been linked with the increase in earnings since 1998 anyway.

  3.45  There is another feature of the gold-plated base case that is not explicit but needs to be addressed. The withdrawal of the rebates for contracting out would not in itself alter the relative attractiveness of final salary (defined benefit) as opposed to money purchase (defined contribution) occupational pensions. However, the ending of the rebates might bring forward the decisions of some companies to end final salary schemes. Whether this is a matter of concern depends on a judgement as to whether final salary schemes are capable of being saved. If you think that the actuarial arithmetic that appears to spell the eventual decline of final salary schemes is remorseless, then a reform that might give that process an extra push might be seen as regrettable but unavoidable.


  3.46  One frequent confusion in the debate over pension or wider welfare reforms is the equating of targeting solely with means testing. The original universal benefits introduced by Beveridge were well-targeted on groups that by and large suffered from low incomes. This included the BSP as in 1945 most pensioners were poor. The fact that by the 1980s and 1990s this was no longer universally the case as many of the retired had good private incomes was one argument advanced for the retreat from the universalism embodied in a somewhat more generous BSP going to all. However, if certain groups of pensioners still suffer from higher levels of poverty, this might justify an increase in the BSP aimed only at these groups. This would be a form of targeted universalism.

  3.47  The easiest way of enabling such targeting would be to increase the BSP for older pensioners who tend to have lower incomes.

Table 3.2


Below 50 per cent of mean
Below 60 per cent of median
Under 70


  After housing costs.


  Households Below Average Income series.

  3.48  This then is the argument for what we have described as the silver plated base case. This consists of raising the BSP to the level of the MIG and indexing it in line with earnings for those pensioners aged over 75. The BSP would remain price indexed for those aged 65-74 and other features of the current pensions settlement would remain in place, including the MIG and the Pension Credit that would still be playing an important role for the 65-74 age group, and the State Second Pension.

  3.49  The available data on the income distribution allowed Hawksworth (2002) to only model an increase in the BSP for those over 75. This would raise direct public expenditure on pensions to 5.8 per cent of GDP in 2010-11 compared with 5.6 per cent in the Government's base case and 6.5 per cent in the gold plated base case with the State Second Pension still in place. In 2050 this option would see direct public expenditure on pensions rising to 7.3 per cent of GDP compared with 6 per cent in the Government's base case and 8.6 per cent in the gold plated base case. Unsurprisingly then the silver plated base case is more affordable than the gold plated base case but costs more than the Government's base case.

  3.50  This reform option scores poorly on the goal of simplifying the system. In fact it adds another layer of complexity with people having to adjust their behaviour to the prospective jump in the BSP at age 75 in the context of a pensions system where all the other features remain intact. This hike in the BSP at age 75 is quite significant. Expressed in terms of benefit levels as they were in 2000, the BSP could be worth just £34 at age 65 in 2050 and would then leap to £92 at age 75. Such a skewed BSP does not seem sustainable.

  3.51  Thus increasing the BSP to the level of the MIG for the over 75s as a more targeted measure to alleviate current pensioner poverty, does not really resolve one of the critical structural issues facing the current pension system: the future of the BSP for everyone. The silver plated base case looks like another sensible pragmatic short-term solution that does not really meet the objectives for a sustainable pensions settlement over the long-term.


  3.52  The fact that the Government's pension settlement now implies an increase in direct public expenditure on pensions allows a range of options other than increasing the BSP to be explored. To many the attractions of going down the route of making the State Second Pension more generous seem obvious. It appears to be more targeted than a straightforward increase in the BSP, having been explicitly designed to be more generous to the lower paid when compared with the SERPS that it is planned to replace.

  3.53  Modelling an increase in the State Second Pension is, however, rather difficult. Hawksworth (2002) assumes that the savings credit rate in the Pension Credit is reduced steadily from 60 per cent in 2010 to zero by 2050. The resources saved are used to increase spending on the State Second Pension so that it costs twice as much by 2050 (that is 2.4 per cent of GDP as against 1.2 per cent of GDP in the Government's base case). To the extent that this extra money is focussed on those poorer pensioners who opt for the State Second Pension, this would lift them above the MIG. This reform option appears to turn out cheaper than the Government's base case, with direct public expenditure running at just 5 per cent of GDP in 2050 as against 6 per cent of GDP. Phasing out the Pension Credit saves some resources and the more generous State Second Pension reduces the cost of the MIG.

  3.54  However, a more generous State Second Pension would need to be accompanied by more generous rebates for opting out otherwise more people will stay in the State Second Pension inflating its direct costs. The extra costs of the rebates cannot be modelled. However, a rule of thumb might be that a doubling of the value of the State Second Pension would need to be accompanied by a doubling of the cost of the rebates. These already cost 0.9 per cent of GDP in 2000-01. This would effectively make the total cost to the exchequer of this reform package similar to the cost of the Government's base case (Hawksworth, 2002).

  3.55  Falkingham et al (2002) also model an increase in the State Second Pension paid for by phasing out the Pension Credit. They point out that any increase in the State Second Pension would have to be considerable for people to avoid means testing in retirement. They also note that the more generous the State Second Pension is the more likely it is that the stakeholder target group will opt back in, thus increasing the direct expenditure costs to the state.

  3.56  This reform package scores poorly against the objective of increasing the clarity of the system, though it does allow the Pension Credit to be phased out eventually. Moreover, the history of SERPS is salutary. It might be relatively easy for a future government to once again reduce the generosity of the State Second Pension safe in the knowledge that few might object because so few understand the system.


  Hawksworth (2002) explicitly scores the four different models against the four key objectives we have identified for the pensions system. The Government base case scores poorly in terms of certainty and transparency when compared with the gold plated case. The gold plated case is nevertheless as affordable as the government's model when combined with the phasing out of the State Second Pension and the associated rebates and the increase in the retirement age to 67 by 2030.

(cost as per cent
of GDP in 2050)
(minimum income
Incentives to save
for own retirement
Certainty and
Index (best =
6,Worst =1)
Government's base case:
PC better for low earners/worse for moderate earners
—without PC
Poor for low earners
Gold plated base case (BSP=MIG for all):
Very high
No means-testing but income effect could be negative
—with no S2P
Similar to above
—retire at 67 from 2030 and no S2P
High for 67+/low for 65-66s
Similar to above
Silver plated base case (BSP=MIG at 75)
Very high for 75+, high for others
Means-testing retained for under-75s
Higher S2P and no PC
Less means testing should boost incentives

  1 PwC Index: Maximum score set at 6 for option 2b as this has no means-testing and no complex second state pension arrangements. For other options, up to three points deducted for the degree of means-testing and 2 points for options where the State Second Pension is retained. An additional 1 point deduction is made where the state retirement age is increased (due to the disruption to pensions planning).

  2 Including the savings from abolition of NIC rebates assumed to be 0.9 per cent of GDP each year.

  3 Excluding the cost of NIC rebates for those opting out of the State Second Pension, which could be significantly higher under this option.


  Hawksworth, 2002.


  4.1  The results from the modelling exercises discussed here lead to the conclusion that it is possible to design an alternative pensions settlement that would better meet the key objectives against which any system should be measured.

  4.2  The current pensions settlement ensures an adequate income for those who claim the MIG and will claim the Pension Credit, but with the drawback of considerable complexity and a very large role for means-testing with the associated problems in terms of take-up, incentives and people's sense of the equity of the system.

  4.3  The gold plated base case, including the phasing out of state second pension provision and a modest raising of the official retirement age to 67 by 2030, best satisfies the key objectives we have set. Setting the Basic State Pension at £100 in 2003-04 prices will (along with the Winter Fuel Payments and other supplements to the Basic Pension) give all those with a full contributions record a non-means tested income at around the "low cost but acceptable" benchmark for the incomes of the elderly. It will be just above the quasi-official poverty line of 60 per cent of median household income after housing costs. It will satisfy the objective of securing adequacy, albeit at a level of income few will regard as overly generous.

  4.4  Importantly, it will do this without the complexities of the existing policy regime. The overall pensions system will be significantly simpler, assuring adequacy without involving significant means testing or a complex and ill-understood state second pension. The proposed increase in the state retirement age to 67 will occur in the decade after the official retirement age for women has increased from 60 to 65, which should make the reform easier for people to comprehend. It thus scores highly in terms of the clarity and transparency of the overall pension settlement.

  4.5  This can be illustrated by showing how far the proposed new settlement would significantly lower the need for potentially costly advice. Referring back to the simple framework outlined earlier for thinking about the different forms of advice needed by individuals:

    —  the question of whether to save should at least be significantly simplified by people having clearer expectations about what the state will provide in a non-means tested manner;

    —  the question of how much to save would be unresolved and is discussed further below in the context of questions in relation to the issue of compulsion;

    —  the question of where to save would be significantly simplified by the phasing out of the State Second Pension and equity based ISAs leaving funded pensions vehicles as the only form of tax-favoured equity based saving;

    —  the question of whether to contract out of the State Second Pension or not would become redundant as the state would withdraw from second pension provision.

  4.6  In these circumstances the questions of who should provide and who should pay for advice would become much less important. Overall the costs of providing advice should decline significantly—a major additional benefit of introducing a clearer and more transparent system.

  4.7  The overall reform package favoured here passes the affordability test as well as the government's pensions settlement, as we have tried to make the package broadly revenue neutral. The overall costs to the exchequer rise to around 6 per cent of GDP by 2050, but this is the same as the projected costs of the Government's settlement including the Pension Credit. Under that settlement the Government is confident that "Over the longer term, the forecast spending on pensions remains affordable and sustainable" (page 7, DWP Nov. 2001). The reform package suggested here is likewise affordable and sustainable.

  4.8  In terms of the impact of the proposed reform package on incentives, the greater clarity and transparency of the system and the reduction in the prevalence of means-testing should give clearer incentives for people to save for their own retirement. However, offsetting this, the more generous Basic State Pension might have what economists would term a negative income effect, that is some people might save less given the extra income guaranteed to them by the state. As the Basic State Pension would still be set at a very modest level and most people would want to retire on a significantly higher income, those with the means would still face a clear incentive to save in a second tier pension vehicle.


Compulsion to save

  5.1  The reform package would of course remove the option of paying into the state second pension and the associated rebates to those contracting out. The second tier vehicles that people could save into would be the various private funded options: occupational, personal and stakeholder pensions. Personal and stakeholder pensions could no longer rely on being fed by the rebates. A key question left unresolved is whether there should be compulsion to contribute to a private funded pension.

  5.2  The reform package suggested by the Pensions Reform Group (Field et al, 2001) gave a clear answer to the question of compulsion. This group advanced a similar analysis of the problems with the current pensions settlement as presented here and suggested that Government policy, with its heavy reliance on means-testing, could be summed up as telling people "don't save, but don't be too sure about relying on us either" (page 38, Field et al 2001). Their proposal for a Universal Protected Pension represented a hybrid of the pay-as-you-go Basic State Pension and a new, state administered (at arms length), strongly redistributive, funded, defined benefit scheme. The overall pension paid out would be much more generous than the enhanced Basic State Pension suggested here. However, this would be paid for by a significant increase in national insurance contribution rates, with these contributions diverted into the funded part of the scheme. This would represent a clear increase in compulsion.

  5.3  In the gold plated reform package outlined here there could effectively be less compulsion, unless individuals were compelled to make minimum payments into some form of private funded pension. People would continue to pay the current contracted-in contribution rates but would lose their rebates for contracting out. Those personal pension pots fed solely by the rebates would now require people to add their own private savings on a voluntary basis if they were not to be effectively frozen in terms of new contributions. This could worry many of the providers of personal and stakeholder pensions, though they now feel obliged not to persuade most people to opt-out and fund their pension pots through the rebates. Some providers would advocate some level of compulsion, in terms of minimum individual contributions into funded pensions, doubting that the greater clarity and transparency of the proposed reforms would in itself allow providers to more effectively market their products. These providers also make the point that where employers make a contribution this can be just the incentive necessary for individuals to take up, say a Stakeholder, and contribute as well.

  5.4  This discussion raises the issue of whether the state should compel employers to make contributions to their employee's pensions, whether they are occupational, personal or stakeholder pensions. Such compulsion does not exist in the current system, though many employers do of course contribute voluntarily. We have, however, seen that the manner and scale of those voluntary employer contributions appears to be changing significantly, with the decline of traditional defined benefit or final salary schemes and a switch to defined contribution schemes often with lower employer contributions. It was emphasised that it was not the switch itself which was the cause of lower employer contributions, rather a range of pressures were causing firms to lower their contribution rates and a switch to a defined contribution scheme was the means to achieve this. It was also emphasised that the true incidence of employer contributions should in an efficient labour market fall on employees in the form of lower gross wages.

  5.5  Nevertheless there may be significant attractions in terms of presentation of there being a new contract between the employer and the employee to match the new contract being suggested here between the state and the individual, with employers and employees both paying into a funded pension. The question is whether this new contract between the employer and the employee should be underpinned by compulsion, as by definition would be the contract between the state and the individual with its more generous taxpayer financed pay-as-you-go Basic State Pension.

  5.6  One problem with compulsory minimum employer and employee contributions into a funded pension is that this can become the norm for contributions rather than the floor, with the potential that current contributions above the minimum could fall towards it. Another objection is that over the lifecycle it is not necessarily optimal for an individual to be making compulsory contributions to a funded pension year-in and year-out, rather than reducing their debts or saving in some other way. One objection in principle to such compulsion is that the State has no right to tell any individual how much and in what ways to save, in the absence of external costs being imposed on the rest of society. One reason why some people advocate greater compulsion in the first place is so individuals do not rely on receiving means-tested benefits in retirement. However, in a pension settlement where the role of means testing is residualised, the need to avoid such "moral hazard" is also minimised.

  5.7  A key argument in relation to compulsion must be that many employers will continue to provide a pensions contribution to their employees as part of the employment contract, albeit inevitably at a lower level of contributions than today. Providers should be able to convince individuals to make contributions, pointing out that although the enhanced Basic State Pension will remove the perceived penalty to saving, it will itself only provide a bare minimum income. The income effect of a higher Basic State Pension in reducing saving is likely to be modest precisely because the Basic State Pension will itself remain quite modest. It is not possible to say whether total private savings will go up or not under this reform package without further compulsion, but then that is not a first order goal for any pensions reform. On balance the proposed pensions settlement may work without additional compulsion, though it is likely that this would be an issue that policy makers would want to come back to at some point in the medium term.

Incentives to save

  5.8  The debate on compulsion is perhaps jumping the gun. Clearly, the alternative strategy to compulsion is to make the existing emphasis on incentives work better for more people. It has been recognised that the current reliance on tax relief as the sole saving incentive does not work for many people, particularly those on low incomes. In their broader savings strategy, the government is pursuing this approach and developing new means of incentivising people to save. We have already referred to the Savings Gateway and the Child Trust Fund, which include the notion of "matched" contributions and kick-start capital endowments respectively. These policies are still being consulted upon so it is early days on how public policy can be more creative in the incentives it offers to get people to save. In the pensions arena, we need to learn form this wider debate and perhaps think more imaginatively about new or reformed incentives, before considering compulsion. A framework of coherent incentives might fit better with British culture and the evolution of our welfare state, and would also avoid some of the problems with compulsion identified above.

  5.9  Assessed against our objectives for pensions reform the proposals of the Pension Reform Group (Field et al, 2001) score highly in terms of adequacy for future pensioners. However, in itself the proposed Universal Protected Pension does not do anything for current poor pensioners. This is why in addition Field et al (2001) suggested that there are significant increases in the Basic State Pension for the over 75s and over 80s—that is a variation of our silver plated base case. Their proposals score highly in terms of the improvement in incentives to save, in so far as the incidence of means testing is reduced, but there is in addition of course the greater compulsion to save anyway. The proposals are clearly affordable in that the reforms are fully funded and the proposed variation on the silver plated base case would not be too costly. However, the proposals do not meet the objective of introducing greater clarity and transparency into the system, as they—intentionally—involve a significant departure from the system as it has evolved over the post-war period and with which people are familiar. A more generous Basic State Pension by its very simplicity would seem more likely to generate sustained popular support than a more wholesale reform that might be hard to explain.

  5.10  Indeed it is the fact that the gold-plated base case builds on the existing settlement and its most widely recognised feature—the Basic State Pension—that may be its strongest point.

Richard Brooks

Sue Regan

Peter Robinson

February 2002

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2002
Prepared 12 April 2002