APPENDIX 25
Letter to Chairman from Government Actuary's
Department (CP 35)
In your letter to Chris Daykin of 10 November,
you asked four questions about the finances of the National Insurance
Fund. I shall answer these in turn and conclude with some clarification
on widow's benefits.
1. ESTIMATE OF
THE NATIONAL
INSURANCE FUND
OVER THE
70 YEARS
BASED ON
CURRENT CONTRIBUTION
RATES
Table 1 shows:
— the estimated benefit expenditure;
— the estimated contribution income
assuming that National Insurance contributions are paid in future
at the rates and on the structure that will apply following the
changes outlined in the March 1999 budget;
— the excess of contribution income
over benefit expenditure;
— the estimated amount of the Fund;
and
— the Fund as a multiple of benefit
expenditure in the year.
At 10-year intervals from 2010-11 to 2060-61.
The figures shown are based on the estimates published in the
recent Quinquennial Review of the National Insurance Fund, and
hence only extend until 2060-61 rather than for 70 years.
Although allowance has been made for the changes
to contributions announced in the March 1999 Budget, the changes
in the Welfare Reform and Pensions Act 1999, the Tax Credits Act
1999 and the introduction of the State Second Pension have not
been allowed for at this stage. The figures for benefit expenditure
are therefore consistent with those shown in table 10.2 of the
Quinquennial Review. The Welfare Reform and Pensions Act reduces
expenditure on Incapacity Benefit and Widows' Benefits. The Tax
Credits Act increases SERPS expenditure. Overall, allowing for
these two Acts would lead to a small fall in benefit expenditure
and a corresponding comparatively small increase in the projected
Fund. The introduction of the State Second Pension increases both
benefit expenditure (including replacing the allowance for home
responsibility protection for SERPS included in the QR) and contracted-out
rebates more significantly and allowing for this would result
in a considerably smaller projected Fund.
The estimates of the balance of the Fund are
shown assuming two different rates of investment return, being
2 per cent above the rate of earnings growth and 3 per cent above
the rate of earnings growth, net of investment expenses and applicable
taxes if any. These are assumptions only, and do not represent
any upper or lower boundaries on possible future investment returns.
As well as depending on general economic conditions, the future
investment return on the Fund will depend on the investment policy
chosen. Currently, the Fund is invested mainly in short-dated
government bonds, which will produce different returns than if
the Fund were invested in equities.
The results are highly sensitive to the assumed
rate of investment return as Table 1 illustrates. In addition,
the long-term estimates of benefit expenditure and contribution
income are themselves subject to considerable uncertainty. Section
8 of the Quinquennial Review shows the effects of varying some
of the assumptions used in the projections of contributions and
benefits. Although this sensitivity analysis has not been repeated
here, it should be borne in mind. In particular, for the first
thirty years the projected contribution income on current rates
exceeds the projected benefit expenditure by about 7 per cent
per annum on average. This is a comparatively small gap and the
outturn could be significantly different, leading to a very different
size of Fund. For the period 2030 to 2060, the picture is slightly
different with the projected contribution income on current rates
exceeding the projected benefit expenditure by increasing amounts
each year and about 30 per cent per annum on average over this
30 year period. This, and significant investment returns on an
increasingly large Fund, drives the growth in the Fund relative
to benefit expenditure over this period.
The introduction of the State Second Pension
is likely to reduce significantly the gap between the contribution
income and the benefit expenditure in the years to 2030 but the
precise effects will depend on the details of the policy finally
adopted and, in particular, the arrangements for contracting out.
The figures shown for benefit expenditure include
administration expenses. The effects of some small items of income
and expenditure in the Fund have been ignored. It has been assumed
that benefit rates and contribution limits will be increased in
line with prices, and that there will be real earnings growth
of 1.5 per cent. Figures are in £ billion in 1999-2000 price
terms.
Table 1
ESTIMATED EXPENDITURE, CONTRIBUTION INCOME
AND FUND FOR THE NATIONAL INSURANCE FUND ASSUMING CURRENT CONTRIBUTION
RATES
| £ billion, 99-00 prices
| 2010-11 | 2020-21
| 2030-31 | 2040-41
| 2050-51 | 2060-61
|
| Contribution income | 60
| 73 | 82
| 94 | 106
| 119 |
| Benefit expenditure | 58
| 66 | 76
| 80 | 80
| 81 |
| Contribution income less benefit expenditure
| 3 | 7
| 6 | 14
| 27 | 38
|
| Investment return of 2 per cent above earnings growth
|
| Fund at end of year | 39
| 114 | 237
| 446 | 889
| 1,652 |
| Fund as a multiple benefit expenditure | 0.7
| 1.7 | 3.1
| 5.6 | 11.2
| 20.3 |
| Investment return of 3 per cent above earnings growth
|
| Fund at end of year | 42
| 128 | 278
| 550 | 1,129
| 2,175 |
| Fund as a multiple of benefit expenditure |
0.7 | 1.9
| 3.6 | 6.9
| 14.2 | 26.7
|
Assuming an investment return of 3 per cent above earnings
growth, the Fund will be of the order of 100 per cent of GDP by
2060-61, assuming GDP increases at a rate of 1.5 per cent above
prices. Thus the Fund will be significant in economic terms and
it is not clear what effect the build up of such a Fund would
have on the wider economy. Depending on how the Fund were invested,
it could have direct implications for investment markets, since
investing a fund of this size could increase the prices of investments
and reduce returns which will have implications for all investors.
For comparison, you may be aware that the assets of occupational
and personal pension schemes are currently of the order of 100
per cent of GDP.
The estimates shown in Table 1 assume that contribution rates
remain at current levels. It is important to consider the purpose
of the National Insurance Fund when studying these results. If
National Insurance benefits are intended to be financed on a pay-as-you-go
basis, then the purpose of the Fund is to act only as a working
balance, so that benefit payments can be met when needed. In this
case, it is likely that future contribution rates would be reduced,
or benefits increased, rather than allowing a large Fund to build
up to the size estimated in Table 1. Alternatively, there may
be an intention to build up a larger Fund than would be required
under a pure pay-as-you-go approach, so that contribution rates
in the future can be smoothed or reduced, or to offer protection
against demographic, or other, changes. It is important to establish
the role and rationale for the Fund if it is to be managed and
controlled in a proper manner. It is also important to consider
who would have control over the Fund and the legal framework in
which they would operate.
6. REVENUE IMPLICATIONS
OF MAINTAINING
BENEFIT ENTITLEMENT
WITHOUT CONTRIBUTIONS
FOR THOSE
BETWEEN THE
LEL AND THE
PRIMARY THRESHOLD
The results shown in this section and in sections 3 and 4
are based on the short-term economic assumptions underlying the
budgetary process rather than the long term assumptions which
underlie the QR projections. This is more appropriate for illustrating
the short term figures for the National Insurance Fund. Results
are only shown for Class 1 contributions, which are contributions
in respect of employees. The split of the effects between the
primary contributions paid by employees and the secondary contributions
paid by employers have not been shown but the figures are available
if required.
Unfortunately, we are unable to isolate the portion of benefit
expenditure in future years which arises from people earning between
the LEL and the primary threshold. However, we have estimated
the reduction in primary (employees') contributions as a result
of raising the starting point for primary National Insurance Contributions
(NICs) from the LEL to the primary threshold. The results are
shown in Table 2, assuming that the primary threshold will be
£76 per week in 2000-01, and then aligned to the single persons
tax allowance from 2001-02 (£88 per week). The figures have
been split between the amount of NICs that goes into the National
Insurance Fund, and the part that is allocated to the National
Health Service. Figures are shown in cash terms, in £ million.
Table 2
EFFECT ON CONTRIBUTION INCOME FROM RAISING STARTING POINT
FOR PRIMARY NICS
| £ million, cash terms | 2000-01
| 2001-02 |
| National Insurance Fund contributions | -740
| -1,550 |
| National Health Service allocation | -100
| -210 |
| Total contributions | -840
| -1,760 |
| Effect as a per centage of total Class 1 receipts
| -1.5 per cent |
-3.2 per cent |
Negative amounts represent a reduction in revenue.
The starting point for secondary (employers') contributions
has been aligned to the single persons tax allowance since April
1999.
7. REVENUE LOSS
FROM THE
FALL IN
THE UEL FROM
136 PER CENT
OF EARNINGS
IN 1982 TO
110 PER CENT
TODAY
Table 3 shows the difference in contribution receipts between
those currently expected, and those which would be expected if
the UEL were set to be 136 per cent of male full-time average
earnings in each year. Figures are given for the years 1998-99
to 2001-02, are shown in £ million in cash terms. The effects
are split between income to the National Insurance Fund, and contributions
that are allocated to the NHS.
The fall in the value of the UEL relative to earnings since
1982 has two effects on contribution receipts. Firstly, gross
(of contracted-out rebates) primary contributions are reduced.
Primary contributions are based on earnings between the LEL and
the UEL. If the UEL falls relative to earnings, then contribution
income is less than it would have been otherwise. Secondly, the
level of contracted-out rebates falls, since rebates are also
based on earnings between the LEL and the UEL. This increases
contribution income, although this second effect is much smaller
than the first one.
The fall in the value of the UEL relative to earnings has
also decreased expenditure on SERPS, since the SERPS pension payable
depends on earnings between the LEL and the UEL in an individual's
working lifetime. The effects of this will be relatively small
in the short term, but would increase over time. This reduction
in SERPS expenditure has not been included in the table below,
but please contact me if you would like to know what the effects
would be.
Table 3
REDUCTION IN CLASS 1 CONTRIBUTION INCOME FROM THE FALL
IN THE UEL RELATIVE TO EARNINGS SINCE 1982
| £ million, cash terms | 1998-99
| 1999-00 | 2000-01
| 2001-02 |
| Actual or expected UEL (£ per week) |
485 | 500
| 535 | 575
|
| UEL set at 136 per cent of earnings (£ per week)
| 570 | 595
| 625 | 650
|
| National Insurance Fund contributions | -690
| -800 | -720
| -580 |
| National Health Service allocation | -130
| -150 | -140
| -110 |
| Total contributions | -820
| -950 | -860
| -700 |
| Effect as a per centage of total Class 1 receipts
| -1.6 per cent |
-1.8 per cent | -1.6 per cent
| -1.3 per cent |
Figures may not sum to totals shown due to rounding.
Negative amounts represent a reduction in revenue.
The pattern of the figures in the above table from 1999-2000
until 2001-02 is a result of the actual UEL being increased in
2000-01 and 2001-02 by significantly larger amounts than normal,
as announced in the March 1999 Budget.
8. REVENUE INCREASE
FROM THE
EQUIVALENT FALL
IN THE
VALUE OF
THE LEL RELATIVE
TO EARNINGS
OVER THE
SAME PERIOD
Table 4 shows the difference between contribution receipts
currently expected and those that would be expected if the LEL
was the same percentage of male full-time average earnings as
it was in 1982—18.4 per cent of these earnings.
Figures are only given for the years 1998-99 and 1999-2000.
From 2000-2001 onwards, the only link between contribution income
and the LEL will be through contracted-out rebates, since the
starting points for both primary and secondary contributions will
be different from the LEL.
I have not included any allowance for the increase in SERPS
expenditure that results from the fall in the LEL relative to
earnings. Nor have I included any allowance for the increase in
expenditure on other benefits that, in principle at least, arises
from the fact that if more people earn over the LEL there will
be more entitlement to (some) benefits.
Table 4
INCREASE IN CLASS 1 CONTRIBUTION INCOME FROM THE FALL
IN THE LEL RELATIVE TO EARNINGS SINCE 1982
| £ million, cash terms | 1998-99
| 1999-00 |
| Actual LEL (£ per week) | 64
| 66 |
| LEL set at 18.4 per cent of earnings (£ per week)
| 77 | 81
|
| National Insurance Fund contributions | +620
| +900 |
| National Health Service allocation | +160
| +180 |
| Total increase in contributions | +780
| +1,080 |
| Effect as a per centage of total Class 1 receipts
| 1.5 per cent |
2.1 per cent |
If the LEL were set to be the same proportion of male full-time
average earnings as it was in 1982, then the LEL would be £81
per week in 1999-2000. This is very close to the starting point
for secondary contributions, which has been aligned with the single
persons tax allowance in 1999-2000, being £83 per week.
Combining the results from tables 3 and 4, the overall effect
of the LEL and the UEL losing value relative to male full-time
average earnings is a reduction in contribution receipts of £40
million in 1998-99, and an increase in contribution receipts of
£130 million in 1999-2000. These amounts are small relative
to the total amount of Class 1 contribution receipts, which are
in excess of £50,000 million in each of these two years.
As mentioned in Section 3 of the Quinquennial Review, if the LEL
and the UEL continue to be increased in line with prices, then
total contribution receipts do not differ greatly from those which
would be received if these limits were increased in line with
average earnings. However, if these limits are increased in line
with prices, the proportion of contributions paid by employees
will decrease over time (since employees' contributions are based
on earnings capped by the UEL) while the proportion of contributions
paid by employers will increase.
9. WIDOWS' SERPS INHERITANCE
I recall that when we gave evidence to the committee there
was some confusion over whether we were discussing the changes
to widows' benefits in the Welfare Reform and Pensions Act 1999
or the problems over the reduction to 50 per cent inheritance
for SERPS benefits due to take effect from April 2000. I hope
that the following information will clarify the position.
1. Total expenditure on widows' benefits (basic and earnings
related benefits combined) is projected to be just under £1
billion per annum up to 2001-02, based on figures from the latest
DSS Departmental Report.
2. The effect on this outgo of the changes in the Welfare
Reform and Pensions Act 1999 were dealt with as the Bill progressed
through Parliament.
3. Total expenditure on the earnings related part of retirement
pension (SERPS) is projected to be just under £4.3billion
in 1999-00 rising to £5.4 billion in 2001-02, based on figures
from the latest DSS Departmental Report. If the reduction to 50
per cent inheritance of SERPS entitlement due from April 2000
is postponed, Fund expenditure will be increased. The increase
will depend on the period of postponement. The written answer
to a PQ reported in Hansard for 25 October 1999 in column 722
gives the estimated cost of postponement. For example, in 1999-00
price terms, the cost is £60 million for 2000-01, £160
million in 2001-02 and £1,160 million in 2010 assuming that
the reduction is postponed until after 2010. These costs are in
respect of those who inherit SERPS whether they are above or below
pension age. In 2010, this extra cost represents about 2 per cent
of benefit expenditure.
I hope that I have answered all of your questions satisfactorily.
Please let me know if you would like any further information.
David Lewis
Chief Actuary
Social Security
2 December 1999
| 
