Memorandum submitted by the Institute
for Public Policy Research (IPPR) (PC 27)
1.1 Since the 1997 election, Labour has
devoted much time and effort to pensions and long-term care reform.
Both areas have generated significant legislation and the commitment
of new public resources. The Government has characterised its
reforms as a new contract for retirement: a settlement for the
1.2 IPPR has great sympathy with the government's
central objective of eliminating pensioner poverty. However, we
have been concerned about many aspects of the Government's reform
package, and retirement policy has not proved to be a source of
political capital for Labour. It has been difficult for Labour
to win public support for its complex mix of pensions policies,
despite significant increases in public spending on pensioners.
1.3 There are several different objectives
that pensions and long-term care policies need to address. These
include: securing adequacy, maintaining affordability, delivering
clarity and simplicity, protecting incentives, and promoting equity.
We do not regard changing the balance between public and private
funding as a first order goal of policy. Not all the objectives
can be satisfied at the same time, and trade-offs between them
are inevitable. By exploring alternative policy options, we have
investigated whether the Government's current settlement should
continue to be pursued or whether a change of direction is necessary.
A NEW CONTRACT
1.4 The Government has made significant
reforms to the pensions environment since 1997. The transition
from SERPS to the new State Second Pension will begin in April
2002 and Stakeholder Pensions have been on sale since April 2001.
The introduction of the Minimum Income Guarantee (MIG), effectively
a big increase in Income Support for pensioners, was a good short-term
mechanism for directing help to the poorest pensioners.
1.5 The Pension Credit, which is now set
to subsume the MIG, also has many positive features: it represents
significant new public funding for lower income pensioners and
it was designed to stop penalising people with small pensions
savings. However, it will require an increase of over 1 per cent
of GDP in pensions spending by 2050, will quickly extend means
testing to the majority of the pensioner population, involves
numerous incentive problems, and represents a further increase
in complexity in an already bewildering pensions environment.
It is very hard to argue that the Pension Credit has been designed
to generate incentives for current workers to save. Rather, it
has been designed to ameliorate the penalties imposed by the MIG
on current pensioners with small second pensions. The only way
in which the Pension Credit represents an incentive to save is
the degree to which it reduces the penalties for future savers
relative to the previous regime.
With low take-up of the MIG, and
take-up certain to be an issue for the Pension Credit, pensioner
poverty is still a key problem.
Incentive problems remain acute.
The Pension Credit does not resolve the incentive problems of
the MIG: instead it spreads them out over a much larger group.
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.
The long-term purpose of the State
Second Pension is unclear. By 2050 anyone with income only from
the Basic State Pension and the State Second Pension will retire
with an income, which is already below the MIG at 65. 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.
Planning is very difficult for individuals.
The pensions environment has become much more complicated since
1997, and the new policy instruments such as the Pension Credit
both increase the need for good quality advice and make such advice
more difficult to provide.
Public support for the pensions settlement
is low. Many individuals feel that means testing retirement benefits
is unfair: the settlement is not seen by the public to be equitable.
1.6 We have compared four sets of pension
The Government's current strategy.
Making the State Second Pension more
Raising the Basic State Pension to
the level of the MIG, and indexing it in line with earnings, for
those aged over 75.
Raising the Basic State Pension to
the level of the MIG, and indexing it in line with earnings, for
1.7 Raising the Basic State Pension to the
level of the MIG, and indexing it in line with earningstogether
with the phasing out of state second pension provision and a modest
raising of the official retirement age to 67 by 2030best
satisfies the key objectives we have set. This option would obviate
the need for the Pension Credit.
1.8 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. 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.
1.9 Importantly, it will do this without
the complexities of the existing policy regime. With the Basic
State Pension raised to a level that would be somewhat higher
than the government's planned level for the combined Basic State
Pension and State Second Pension, the opportunity would arise
of simply phasing out the latter altogether. The need for the
Pension Credit would disappear. 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 would 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. The proposed new settlement would significantly reduce
the need for potentially costly pensions advice.
1.10 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. Raising the retirement age to 67 combined with
ending state second pension provision and with it the system of
National Insurance rebates achieves this goal. 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" (DWP, November
2001). The reform package suggested here is likewise affordable
1.11 Raising the official retirement age
needs to be justified on more than just the grounds of improving
the affordability of the state pensions system. Average life expectancy
has been rising for many years, and for those who will reach 65
in 2030 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. There is evidence that raising
the official retirement age has a positive impact on labour force
participation rates, achieving one of the goals that everyone
recognises as desirable.
1.12 Our favoured reform package might raise
concerns 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. The distributional consequences
of these reforms could be offset through changes to the current
tax treatment of pensions and other forms of saving. The tax system
could be made more progressive by abolishing equity ISAs that
will from 2004 only be benefiting higher rate taxpayers and by
scaling back the age related income tax allowances.
1.13 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 BSP might
result in some people saving less given the extra income guaranteed
to them by the state. As the BSP 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. The proposed pensions
settlement may work without additional compulsion, though it is
likely that policy makers would want to keep this issue under
1.14 IPPR has investigated whether the Government's
current pensions settlement should continue to be pursued or whether
a change of direction is necessary. We have firmly concluded the
latter. Although we are supportive of the Government's aims and
its primary focus on tackling pensioner poverty, we part company
on how best to achieve those aims. If a means-tested approach
was working for poor pensioners and working for people planning
for retirement, then we would be in accord with the Government.
But it is not working, and people perceive means testing in retirement
to be different from means testing benefits during working life.
Even if a reinvigorated campaign saw take-up of the MIG/Pension
Credit improve, many pensioners would still be reliant solely
on the Basic State Pension, which if price indexed will fall further
and further below the poverty line. The introduction of the Pension
Credit, complexity aside, creates incentive problems that cannot
be resolved. In theory, the government's strategy has merit. In
practice, it falls short of eliminating pensioner poverty and
providing an environment whereby people can understand their entitlements,
save and be rewarded for doing so.
1.15 The centrepiece of our suggested reformsto
increase the Basic State Pension to the level of the MIG and ensure
it retains its value in relation to earnings in the futureis
the only way to guarantee all pensioners escape poverty. Moreover,
it allows a radical simplification of the systemthe Pension
Credit becomes redundant and SERPS/ State Second Pension can be
closedcreating a framework which people can understand.
This policy reform delivers for both current pensioners and people
currently planning for retirement. It delivers for people in poverty
and it delivers for the working poor who will no longer be penalised
for their thrift. We also believe it delivers for government:
with the closing of SERPS/State Second Pension and a modest increase
in the official retirement age to 67 it is affordable, and the
political rewards of what would be a highly popular reform would
1.16 Ensuring the security of all citizens
in retirement is a key element of social justice. If the elderly
in our society are in poverty or receiving inadequate care, we
are failing to make progress. We hope our recommendations describe
a contract, which will ensure better provision of pensions for
millions of older people, both now and in the long term future.
2. THE MINIMUM
The Minimum Income Guarantee
2.1 In April 1999 MIG replaced Income Support
for people who are over 60. Like Income Support, MIG is a means
tested entitlement. People qualify on the basis of having low
income and small savings, but the value of their home is not taken
into account. In the 1998 Green Paper the Government stated its
long-term ambition, but made no commitment, to raise the MIG in
line with earnings (DSS, 1998).
2.1 THE MINIMUM
The full weekly MIG rate in 2001 is £92.15
for a single person and £140.55 for a couple, equivalent
to just under 20 per cent of average male full time earnings or
just under 30 per cent for a couple. The income and savings of
a couple are added together for the means test. Most income is
fully set off against any entitlement, so a single person with
a full BSP of £72.50 and no other income or savings would
be entitled to £19.65 from the MIG, bringing their total
income up to £92.15 per week. Savings of up to £6,000
are not taken into account in the means test. People with savings
of between £6,000 and £12,000 receive less help the
more savings they have, and individuals or couples with over £12,000
of savings do not qualify for any MIG payment.
2.2 The MIG is currently at the level which
the BSP would have attained if the earnings link had not been
broken in 1981. The MIG will be raised to £100 a week for
a single person by April 2003 (HMT 2000), and the Chancellor stated
that the MIG would rise in line with earnings for the duration
of the current Parliament (PBR, 2001). The MIG would thus be just
above the quasi-official poverty line
from 2003, so effectively abolishing pensioner poverty if there
were 100 per cent take-up.
2.3 The following table details the Department
of Social Security's own estimates of take up for Income Support
(Minimum Income Guarantee), Housing Benefit and Council Tax Benefit
among pensioner households (Department of Social Security, 2001).
INCOME RELATED BENEFITS TAKE UP AMONG PENSIONER
| ||Pensioner households entitled but not receiving benefit
||Median weekly amounts unclaimed
||Mean weekly amounts unclaimed
|1998-99||19-32 per cent
|1999-2000 (MIG)||22-36 per cent
|1998-99||4-11 per cent
|1999-2000||7-15 per cent
|Council Tax Benefit|
|1998-99||26-32 per cent
|1999-2000||30-36 per cent
Department of Social Security, Income Related Benefits Estimates
of Take-up in 1999-2000.
2.4 As the table indicates, between optimistically one
in five and pessimistically one in three pensioner households
did not claim the Minimum Income Guarantee payments they were
entitled to in 1999-2000.
Whilst some of these people would have been missing out on very
small amounts, the median and average weekly amounts unclaimed
are significant: £12.80 and £20.00 respectively. What
makes these take up figures so important is that all those entitled
but not receiving this benefit are, by definition, living in poverty.
2.5 It is notable that pensioner take up of Housing Benefit
is much better than take up of the Minimum Income Guarantee. This
may be because many pensioners claiming HB were already claiming
before they became pensioners: non pensioners and pensioners have
similar take up rates for HB, whilst take up of MIG is much worse
than take up of Income Support for younger households.
2.6 IPPR has conducted its own research into people's
attitudes towards means testing with a series of focus groups
and interviews with a spectrum of age and income groups. The results
reveal a complex and sometimes contradictory picture (Edwards,
2001). Firstly, few people instinctively see means testing as
a way of targeting help on those who are most in need. Instead,
the first reaction of many is to see means testing as a way of
penalising those who have been hard working and prudent. There
is a great deal of resentment towards a perceived group of the
"undeserving poor" who are felt to receive help they
do not deserve. The concept of the undeserving poor is also mirrored
by a perception of the "deserving rich": those people
who have worked hard and deserve their relative wealth.
2.7 However, when talking about these undeserving poor
our interviewees were rarely referring to people they knew. People
who "abuse the system" and get "handouts"
they don't deserve are most often an unknown quantity and more
likely to be shaped by stories in the media than by individual
experience. When means-testing is related to known individuals
such as friends, family, and neighbours, people are often far
more willing to make the arguments in favour and to recognise
a genuine need. Our research revealed a real concern for pensioners
struggling to get by on very limited resources.
2.8 In some cases the same people who object to the "undeserving
poor" receiving special help, simultaneously object to what
they perceive as the stigma attached to means testing, the difficulty
of claiming entitlements, and the intrusion of privacy implied
by the process. Another area of particularly stark contradiction
was the conflict between people's desire for the flat rate benefits
and rewards according to contribution. Some responses suggested
that this contradiction could be resolved by a decent flat rate
pension for all with additional provision being on a contribution
related basis. Others expressed the view that people should "get
out what they put in", and few recognised that the state
should have an essentially redistributive role.
2.9 People often feel particularly negative towards means-testing
when it relates to older people rather than those of working age.
At the broadest level there is an argument that older people are
more vulnerable, the implication being that older generations
are a homogenous group who are collectively in need of support.
Other, perhaps more fundamental, arguments are made around the
notion that the wealth and assets accrued in retirement have been
built up over the course of a life. There is a perception that
retirement wealth is the result of years of hard work, saving
and good money management. Many therefore resent these assets
being "held against" people as it suggests penalising
them for the effort they have put in over the years. In general
our interviewees found it much more natural to consider as "fair"
treating retired people alike than recognising inequalities and
responding to them by treating people differently.
2.10 Concerns quickly arose that the Minimum Income Guarantee
might undermine incentives as people who worked and accumulated
pension entitlements found themselves with an income in retirement
little higher than the MIG. The MIG effectively operates a 100
per cent rate of benefit withdrawal as original income rises up
to its level, and this has been felt most acutely by those with
small second pensions. For example, a single pensioner with a
full BSP plus a weekly pension of £15 would be no better
off than a pensioner with only the BSP and no additional income,
assuming both claimed their full MIG allowance. The saved income
of £15 would be fully set off against MIG entitlement. This
problem was given real bite by the large number of pensioners
in this situation: £15 was in fact the average SERPS payment
to men aged 65 to 69 in 1998,
and the figure was just £8 for women in the same age group.
Around a third of all pensioner benefit units received less than
£5 a week from private sources in 1995-96 (Emmerson and Johnson,
2.11 In addition to the incentive problems, the very
high withdrawal rate operated by the MIG generated public resentment
over the perceived unfairness of the system. It was clear that
both the incentive and fairness problems would become progressively
more acute if the BSP were to grow in line with prices and the
MIG with earnings. The response to this problem was the Pension
Credit announced in the autumn of 2000.
2.12 The Pension Credit has been proposed for introduction
in 2003, and will subsume the Minimum Income Guarantee. Those
on low or modest incomes will have their income raised to a guaranteed
minimum, and will then receive a further "savings credit"
on any other income they have in excess of the level of the Basic
State Pension up to a ceiling. The savings credit then reduces
in value as the individual's original income increases to a second
ceiling, above which no credit is payable. The guaranteed minimum
can be considered the replacement for the MIG. The savings credit
is the element, which is intended to deal with the incentive problem
of the MIG and reward those with small retirement incomes additional
to the BSP. The guarantee element operates for men and women over
the age of 60, and the savings credit for all pensioners over
the age of 65. The future uprating of both the guarantee element
and the lower savings credit threshold are crucial aspects of
the future development of the Pension Credit. Whilst the Government
has announced that the guarantee element is set to rise in line
with earnings at least until the end of this Parliament, there
is no commitment beyond that date (Department for Work and Pensions,
BOX 2.2 THE
At the time of introduction in 2003 the guaranteed minimum
income is expected to be £100 per week for a single pensioner
and the Basic State Pension £77 per week. Income excluding
any Pension Credit, called original income, between £77 and
£100 will attract a credit of 60 pence per £1. Thus
an individual with a full BSP plus £10 of second pension
would have an original income of £87 and would receive £100
- £87 = £13 in guaranteed income top up plus £10
x 0.60 = £6 as savings credit. The maximum savings credit
payable would thus be £23 x 0.60 = £13.80 for someone
with an original income of £100. The savings credit is then
tapered off at the rate of 40 pence per £1 of additional
original income, so an individual with original income of £101
receives £13.40 of savings credit, an individual with original
income of £102 receives £13.00 of savings credit, and
someone with £135 of original income receives no Pension
Credit. Analagous arrangements will operate for couples.
The Pension Credit will also treat savings more generously
than the MIG. As with the MIG, savings of up to £6,000 are
not taken into account in the means test for the Pension Credit.
However, the upper savings limit of £12,000, above which
people are not entitled to the MIG, will not apply for the Pension
Credit. Instead, savings above £6,000 will attract notional
interest at 10 per cent which will be taken into account as if
it were income. This notional interest rate is half that assumed
under the MIG.
An additional improvement to the MIG is that entitlements
to Pension Credit payments will be reassessed relatively infrequently:
initially at retirement, then at 65, then only once every five
2.13 The most obvious way of considering the Pension
Credit is as a straightforward means tested benefit with a long
taper. The maximum benefit for an individual with a full state
pension is £23 if they have no additional income. For each
£1 that their income rises above this level, the £23
is reduced by 40 pence, a taper which runs up to those earning
(£77 + £23/0.4) = £134.5. The problem with the
MIG was that it had a 100 per cent withdrawal rate up to a ceiling
of £100 (in 2003 terms). The problem with the Pension Credit
is that it has a withdrawal rate of 40 per cent up to a much higher
ceiling of £135, therefore affecting a much greater number
2.14 The Pension Credit will generate other incentive
problems. The first is that those with entitlements to less than
the full level of the BSP will face a 100 per cent withdrawal
rate on any additional earnings up to the full BSP level. As noted
above, over 10 per cent of men were in this position in 1997 (Pension
Provision Group, 1998). Individuals in this position will have
no incentive to save if the expected weekly income from their
saving will not take them over the level of the BSP.
2.15 The second incentive problem arises because of the
gap between entitlement to the guarantee element of the Pension
Credit at the age of 60, and entitlement to the savings credit
at the age of 65. Women pensioners between the ages of 60 and
65 will thus remain in the same position as under the MIG with
respect to benefit withdrawal as a result of small second pension
incomes. Poorer female pensioners may see no benefit from saving
for small second pensions until they reach the age of 65.
2.16 The third incentive problem is that those with original
incomes between £77 and £100 face new and conflicting
incentives. Their withdrawal rate has fallen from 100 per cent
to 40 per cent, improving their incentives to save. On the other
hand the effect of this is to increase their total income, which
may reduce their need to save. The net effect of these changes
is not easily predictable. However, the 100 per cent withdrawal
rate in this income band was clearly politically unacceptable,
and its removal helps to address the popular concern that the
MIG was actually unfair.
2.17 The fourth incentive problem is that original income
in the range £100-£135 now faces an increased withdrawal
rate. Under the MIG these individuals suffered 100 per cent withdrawal
on income below £100 and zero withdrawal above that level.
They would now face 40 per cent withdrawal rates between £77
and £135, reducing their incentive to save into the £100
to £135 bracket. In addition, the Pension Credit will increase
their total incomes, reducing the need to save. The Pension Provision
Group noted in their response to the consultation paper that "it
seems likely that saving which would result in income in this
range (£100 to £135) would fall" (Pension Provision
2.18 The fifth incentive problem concerns the interaction
of the Pension Credit with Housing Benefit and Council Tax Benefit.
All those receiving MIG are currently entitled to the maximum
HB and CTB. Income above MIG level reduces entitlement to HB by
65 pence per £1, and entitlement to CTB by 20 pence per £1:
a maximum 85 per cent withdrawal rate. Under the Pension Credit
income up to the level of the guaranteed minimum plus the maximum
savings credit (£113.80) is likely to be disregarded when
calculating entitlement to HB and CTB.
This would mean that each £1 of additional income above £113.80
would incur 40 pence per £1 withdrawal of Pension Credit
plus £0.60 x (0.65+0.20) = 51 pence per £1 withdrawal
of HB and CTB. It is hard to identify any treatment of the three
benefits, which does not impose very high withdrawal rates at
some point in the income distribution. On the other hand it should
be remembered that under the existing system people claiming HB
and CTB already face 85 per cent benefit withdrawal rates on some
of their income above the MIG level.
2.19 Finally, incentives need to be understood to be
effective. The foregoing discussion should make it clear that
the Pension Credit is a highly opaque means of generating the
right incentives: it is extremely difficult for individuals to
know what to expect from it, especially over the long time periods
relevant to retirement planning. This problem does not just affect
the individual planning for their own retirement: professional
advisers are in a similar position in respect of their clients.
2.20 It would be very hard to argue that the Pension
Credit has been designed to generate incentives for current workers
to save. Rather, it has been designed to ameliorate the penalties
imposed by the MIG on current pensioners with small second pensions.
The only way in which the Pension Credit represents an incentive
to save is the degree to which it reduces the penalties for future
savers relative to the previous regime. The story of the Pension
Credit highlights one of the fundamental problems of pensions
policy: the requirement to meet the needs of both current and
2.21 The proposals before Parliament at the beginning
of 2002 indicated short term costs of £2 billion for 2004-05,
including £460 million for consequent increases in Housing
Benefit and Council Tax Benefit payments. It has been independently
estimated that the Pension Credit will increase public spending
on pensions alone by about 1.2 per cent of GDP by 2050 (Hawksworth,
2002). The Department
for Work and Pensions produced its own long term estimates for
the cost of the Pension Credit in late January 2002 and these
indicate similar incremental expenditure of 1.3 per cent of GDP
by 2050 (Department
for Work and Pensions, 2002).
On the basis of the independent modelling, total government spending
on pensions and related income support is thus set to rise from
just under 5 per cent of GDP to 6 per cent by 2050, in contrast
to all previous projections which had public spending falling
as a proportion of GDP. One of the key aims of the 1998 Green
Paper would appear to no longer apply.
2.22 It has been estimated that the Pension Credit will
increase the proportion of people being means tested to approximately
65 per cent of single pensioners and 51 per cent of couples by
2025 (ABI 2001). The Department for Work and Pensions also estimates
that coverage will be 65 per cent of pensioner benefit units by
2040 if the guarantee element is uprated in line with average
earnings whilst the savings credit threshold rises in line with
prices (Department for Work and Pensions, 2002). The significant
increase in means testing to cover the majority of the retired
population appears to undermine one of the other main aims of
the 1998 Green Paper. It is possible that developments in other
areas of public policy, particularly in-work family benefits and
tax credits will mean that future cohorts may increasingly be
used to receiving means tested benefits. However, it remains a
challenge to ensure maximum take up and minimum effects on work
and saving incentives. However, there are significant differences
in which in work benefits are perceived compared to benefits in
retirement. New benefits such as the Working Families' Tax Credit
are received thorough the pay packet and are effectively seen
as a supplement rather as an entitlement. IPPR's qualitative work
brings out that the public do see means testing in retirement
as different from means testing in-work benefits: whilst those
receiving benefits in work may be getting something "extra",
those who "fail the means test" in retirement are often
seen as being punished for their accumulated work and saving (Edwards
et al, 2001).
2.23 The Pension Credit raises other questions. Its introduction
into an already complex system of provision will add to the lack
of transparency, making it harder still for individuals to understand
that system. In the long-term, it is not clear that the Pension
Credit and the State Second Pension will work together. The ostensible
aim of the State Second Pension added to the Basic State Pension
is to provide a non-means tested retirement income for those on
low lifetime incomes that is clear of the means-tested MIG level.
There were already concerns before the introduction of the Pension
Credit that the State Second Pension would not achieve this aim
without being made significantly more generous (Rake, Falkingham
and Evans, 1999). With the introduction of the Pension Credit,
the State Second Pension and all other forms of second pension
provision will have to work much harder to give people a non-means
tested retirement (Pension Provision Group 2001).
2.24 Not withstanding what has just been said, it is
still important to recognise that the Pension Credit represents
an improvement on the Minimum Income Guarantee. It undeniably
represents significant new public funding for lower income pensioners,
and it softens some of the benefit withdrawal problems of the
system it replaces. However, it is clear that it is far from a
perfect solution, and that we must ask the question whether the
very significant resources it will require could be better deployed.
2.25 Direct public spending on pensions in 2000-01 totalled
five per cent of GDP. This included spending on the Basic State
Pension, SERPS and means tested benefits including income support
or the MIG and housing and council tax benefit. However, the Government
also contributes resources to people's pensions through its treatment
of pensions through the tax system and its system of National
Insurance rebates for those who contract out of state second pension
STATE RESOURCES SPENT ON THE PROVISION OF RETIREMENT INCOME,
| ||Percentage of GDP
|Direct spending items
|Basic State Pension||3.6
|Other (Winter fuel allowances etc.)
|"Expenditure" on "real" tax concessions
|Contracted out NI rebates||0.9
|Tax free lump sum||0.2
|Total expenditure on "state" support
|Income tax relief for approved pension schemes
|(Of which: higher rate tax relief||0.2
|Income tax relief on PEPs/ISAs/TESSAs||0.2
|Upper earnings limit on NICs||0.6
HM Treasury 2001, IFS
2.26 The system of NI rebates is especially costlya
sum equivalent to 0.9 per cent of GDP was paid into people's pension
pots in 2000-01 in return for their giving up their claims to
state second pension provision. This was almost twice as much
as was paid out in SERPS itself in that year. This system of NI
rebates is almost unique to the UK, with only Japan also allowing
people to contract out of the state pension system but on a much
smaller scale (Disney and Johnson, 2001, p25). The introduction
of rebates for personal pensions in the late 1980s cost the government
substantial revenues that it will never recoup from lower SERPS
2.27 Individual contributions into pension schemes attract
income tax relief at an individual's marginal rate. There are
limits on the proportion of annual income that can be saved in
a pension and these are age-related in the case of personal pensions.
Income and capital gains accrued in pension funds are tax-relieved.
Pension income is then taxed when it is paid out. This approximates
the treatment that would exist under what economists call an expenditure
tax regimeincome is tax-free on the way in but taxed on
the way out (Emmerson and Johnson, 2001).
2.28 However, other features of the tax treatment of
pensions make them especially tax-favoured. Two features stand
out that benefit existing pensioners. Firstly, income tax allowances
are higher for pensioners. In 2001-02 the personal allowance for
those aged under 65 was £4,535. It was £5,990 for those
aged between 65-74 and £6,260 for those aged 75 and over.
Pensioner couples still benefit from a married couple's allowance
when this has been abolished for all those under 65. Secondly,
a quarter of an individual's pensions pot or 1.5 times final salary
in the case of a salary related defined benefit scheme can be
taken as a tax-free lump sum.
Together these two forms of "tax expenditure" currently
cost the Exchequer about two-fifths of 1 per cent of GDP. In addition
employer contributions to pensions are exempt from NICs and no
NICs are paid by individuals when the pension is drawn.
2.29 Although the whole cost of the income tax relief
allowed for pension contributions should not be counted as an
"expenditure" in that people will be paying tax on their
pension income, some pensioners will be better off in that they
may get tax relief at the 40 per cent tax rate while in work but
attract only the basic rate of tax in retirement. This has made
higher rate tax relief a target for revenue raising. Overall the
system of tax relief for pensions is highly regressive.
2.30 Overall expenditure on the "real" tax
concessions and subsidies in the pensions system totalled about
1.3 per cent of GDP in 2000-01. This needs to be added to the
cost of direct public spending on pensions. In total the state
devoted 6.3 per cent of GDP to pensions in 2000-01. In looking
at the scope for reforms this is the total pot that one would
look at if an attempt was being made to make any set of reforms
revenue neutral, that is to ensure that the overall costs to the
state were similar to those envisaged under the Government's pensions
2.31 There are other features of the tax system relevant
to the debates on pensions and long-term care. One anomaly in
the tax system that has been the subject of much debate is the
upper earnings limit on national insurance contributions. Employees
in 2001-02 paid NICs at 10 per cent on income between £87
and £575 per week, but not above this upper limit, producing
a dip in the marginal direct tax rate from 32 per cent to 22 per
cent until the higher rate of income tax at 40 per cent clicked
in at £33, 935. Eliminating this limit entirely would have
raised revenue equal to about 0.6 per cent of GDP in 2000-01 and
aligning it to the higher income tax rate would have raised 0.1
per cent of GDP.
2.32 Pensions are not the only form of saving that is
currently tax-favoured. In 2000-01 the tax reliefs attached to
PEPs, ISAs and TESSAs cost the Exchequer about 0.2 per cent of
GDP. There is a very real question as to why the state feels it
should encourage saving through these kinds of vehicles. In the
2001 Budget, the Government outlined three reasons for helping
people to save: for independence throughout their lives; for security
if things go wrong; and for comfort in old age. Private pensions
help secure the first and third of these objectives. Very liquid
assets such as cash ISAs are appropriate for security. The proposed
asset based welfare vehicles are designed to secure the first
objective. What objectives do equity based ISAs meet?
2.33 Equity based ISAs also create real incentive problems
for people thinking about locking away income for up to 40 years
in their pension. Why should one do so when putting income into
an ISA attracts somewhat similar tax treatment as saving in a
pension, albeit on the basis of income not being tax-free on the
way in but tax-free on the way out?
2.34 From 2004 the advantages of investing in an equity-based
ISA will be zero for a basic rate taxpayer due to the little noticed
abolition of the 10 per cent tax credit on dividend income: it
will become a vehicle solely of benefit for higher rate taxpayers.
This will make the continuation of equity-based ISAs very difficult
for a Labour government to justify. The problem of investment
in ISAs competing with investment in pensions would be partially
removed if equity-based ISAs were simply abolished and cash-based
ISAs retained with a modest limit on total savings to help people
shelter some of their "rainy-day" savings in a tax favoured
3. ANALYSIS AND
3.1 Is it possible to design an alternative settlement
that would better meet the key objectives that we think any pensions
system should be measured against? These objectives are:
Ensuring that all receive an income in retirement
that is at least adequate to lift them clear of the poverty line.
Constructing a system that is clear enough for
people to understand its main features so they can plan with some
confidence for their retirement needs.
Adequately rewarding and incentivising people
to save for themselves.
not setting up future costs for the state that
are unaffordable in the sense that they imply a sharp and ongoing
rise in spending as a proportion of GDP.
3.2 We believe that a pensions settlement that could
meet these objectives would be more likely to generate widespread
popular support and be more politically robust. We do not believe
that the current settlement is sustainable. We also believe that
the clarity and transparency of any settlement needs to become
a first order goal of any reform.
3.3 Any discussion of the options for pension reform
has to start with a full understanding of how the pensions settlement
proposed by the Government will look once the system, with its
many separate features, is fully in place. This will then provide
us with a benchmark against which to assess other reform options.
3.4 Our discussion of reform options draws extensively
on two pieces of modelling work that have been done in collaboration
with IPPR. The first modelling exercise has been undertaken by
John Hawksworth at PricewaterhouseCoopers (Hawksworth, 2002).
This has attempted to cost out different pension reform options,
projecting the costs as a proportion of GDP for each decade to
2050, and assessing the options against the four objectives outlined
above. The second modelling exercise has been undertaken by a
team at the ESRC SAGE Research Group (Falkingham, Rake and Paxton,
2002). This has attempted to assess different pension reform options
according to how well they deliver to individual women and men
at the point of retirement and through later life. It should be
noted that IPPR's policy conclusions are not the responsibility
of these external partners.
3.5 These modelling exercises have looked at four possible
sets of reforms. However, here we concentrate heavily on a comparison
of the Government's base case with what we describe as the gold
plated base case.
3.6 The Government's base case The starting point is
the Government's own pensions settlement, what we have called
the Government's base case. This involves a price-indexed Basic
State Pension, an earning-indexed MIG and the introduction of
the State Second Pension and the Pension Credit. One can explore
variations on this base case including for example not introducing
the Pension Credit.
3.7 The gold plated base case This is one of the most
generous settlements that we might envisage. It involves raising
the Basic State Pension gradually to the level of the MIG between
2002-03 and 2010-11 and then indexing it in line with earnings.
In this option it is assumed that the Pension Credit is not introduced
(though it is also compatible with the Pension Credit only being
introduced as an interim measure and then phased out by 2010-11).
In this case the MIG retreats to being a residual means tested
benefit for those with incomplete entitlements to the Basic State
Pension. One can also explore variations on this option such as
ending state second pension provision or increasing the official
3.8 The silver plated base case This option involves
raising the Basic State Pension to the level of the MIG at age
75 reflecting the higher levels of poverty faced especially by
older single pensioners. This can be modelled with a price indexed
Basic State Pension and the Pension Credit continuing for younger
pensioners or with an earnings indexed BSP and no Pension Credit
for younger pensioners.
3.9 Making the state second pension more generous This
option involves putting significant extra resources into the State
Second Pension, which would allow the Pension Credit to be phased
out over the very long-run and reduce the need for the MIG to
top-up incomes. This can also be modelled with different assumptions,
particularly about the level of the BSP.
3.10 These four options are of course not exhaustive.
Even more generous than what we have called the gold-plated base
case would be the introduction of a citizen's pension, where everyone
would receive something equivalent to the BSP as a right without
having to establish a contributions record. If it is correct that
by 2020 most women and men will have achieved full entitlement
to the BSP in their own right, then the current system will look
more like a citizen's pension. However, the current system where
people receive a benefit based on some record of contributions
appears to be well understood and well supported. Given this our
objective of promoting clarity and transparency would seem to
suggest not overturning this feature of the current settlement.
3.11 The least generous option would be for the state
not only to withdraw from second tier pension provision but to
withdraw from first tier provision as well and retain only a means
tested system to help the poorest in retirement, an option modelled
by Falkingham et al (2002). A less drastic option would
be to incentivise individuals to opt out of the BSP in the same
way that they were encouraged to opt out of SERPS into personal
pensions in the 1980s. In some discussion this option is based
on an argument about the inherent superiority of funding over
pay-as-you-go, an argument we do not find compelling. The impact
on the state's finances would, if the system were actuarially
fair, be neutral in the long run: the state would lose income
in the short run as it paid out rebates and gain in the long run
as fewer people had entitlement to the BSP. If the incentives
to opt-out were made too generous the settlement would prove less
affordable for the state in the same way as the reforms in the
late 1980s generated net losses for the exchequer. It would introduce
another element of complexity into the pensions system and would
not reduce either existing or future pensioner poverty. It therefore
fails to meet any of the objectives we have set for pension reform.
3.12 The four pension settlements we explore here encompass
the likely range of options that most people on the centre-left
and probably the centre-right would be comfortable with. They
allow us to probe in more detail the trade-offs inherent in the
objectives we are trying to meet for the pensions system. We think
the analysis leads us to choose a variation on the second optionthe
gold plated base caseas the best outcome, but exploring
the other options enables us to clarify the case for the option
we have chosen.
ESTIMATES OF TOTAL DIRECT SPENDING ON PENSIONS AS PERCENTAGE
|Government's Base Case:|
|Gold Plated Base Case:|
|BSP=MIG for all, no PC||5.0
|& close SERPS/S2P||5.0
|& abolish rebates||5.0
|& raise ret. age to 67||5.0
|Silver Plated Base Case:|
|BSP=MIG at 75||5.0
|Higher State Second Pension:
|without PC (not including
increased costs of rebates)
Hawksworth 2002. Authors calculations on impact of abolishing
That is 60 per cent of median household income after housing
The estimates are constructed by comparing benefit claims with
responses to the Family resources Survey. Ranges represent 95
per cent confidence intervals. Back
The figures in the table above should not be interpreted as showing
a statistically significant worsening of take-up rates between
1998-99 and 1999-2000. Take up rates are estimates given as 95
per cent confidence intervals, and in all cases shown the confidence
intervals for 1998-99 overlap with those for 1999-2000. Back
This is the average for those with SERPS entitlements, who comprised
91 per cent of men in this age category in 1998. Back
These details have not yet been finalised by DWP. However, setting
the disregard at the guaranteed minimum level of £100 would
mean that income between £100 and £135 would attract
a withdrawal rate of 40 pence per £1 for the Pension Credit
and then an additional £1 x (0.65+ 0.20) = 85 pence per £1
for HB and CTB. This would result in a potential withdrawal rate
of 125 per cent between these income limits: clearly unacceptable. Back
This does not include knock-on effects to Housing Benefit or
Council Tax Benefit. Back
On the assumption that the guarantee element of the Pension Credit
is uprated in line with average earnings in all years after 2003
and the savings credit threshold in line with prices. Back
The two exercises use somewhat different modelling techniques
and assumptions which are discussed in Hawksworth, 2002. Back
Also includes free TV licences for the over 75s, widow's pensions,
war pensions and the Christmas bonus. Back
On the condition that the member has at least 10 years of service. Back