OCCUPATIONAL AND PERSONAL PENSION SCHEMES
REVIEW OF CERTAIN
CONTRACTING-OUT TERMS
Reports by the Government Actuary and the Secretary of State for Work
and Pensions in accordance with Sections 42(1), 42B(1) and 45A(1) of the
Pension Schemes Act 1993
Presented to Parliament pursuant to Sections 42(1), 42B(1) and 45A(1) of
the Pension Schemes Act 1993 and Social Security (Reduced Rates of
Class 1 Contributions, Rebates and Minimum Contributions) Order 2011
February 2011
OCCUPATIONAL AND PERSONAL PENSION SCHEMES
REVIEW OF CERTAIN
CONTRACTING-OUT TERMS
Reports by the Government Actuary and the Secretary of State for Work
and Pensions in accordance with Sections 42(1), 42B(1) and 45A(1) of the
Pension Schemes Act 1993
Presented to Parliament pursuant to Sections 42(1), 42B(1) and 45A(1) of
the Pension Schemes Act 1993 and Social Security (Reduced Rates of
Class 1 Contributions, Rebates and Minimum Contributions) Order 2011
February 2011 London: The Stationery Office £
20.50
© Crown copyright 2011
You may re-use this information (not including logos) free of charge in any format or medium,
under the terms of the Open Government Licence. To view this licence, visit
http://www.nationalarchives.gov.uk/doc/open-government-licence/ or write to the Information
Policy Team, The National Archives, Kew, London
TW9 4DU, or e-mail: [email protected].
Any enquiries regarding this publication should be sent to:
Joanne Meusz
Government Actuary’s Department
Email: [email protected]
Richard Jordan
Department of Work and Pensions
Email: richard.jordan1@dwp.gsi.gov.uk
This publication is also available on http://www.official-documents.gov.uk/
ISBN: 9780108510090
Printed in the UK by The Stationery Office Limited
on behalf of the Controller of Her Majesty’s Stationery Office
ID: 2412531 02/11 8544 19585
Printed on paper containing 75% recycled fibre content minimum.
Table of Contents
Review of rebates and reduced rates of National Insurance contributions
for members of defined benefit contracted-out pension schemes
1 Overview .................................................................................................................... 1
2 Introduction to contracting out .................................................................................... 4
3 The 2012 – 2017 defined benefit contracted-out rebate review ................................. 6
4 Factors affecting the cost of providing contracted-out benefits .................................. 8
5 Deriving the defined benefit contracted-out rebate .................................................. 12
6 Valuation approaches / assumption sets .................................................................. 14
7 Demographic assumptions ....................................................................................... 17
8 Financial assumptions – asset allocation ................................................................. 21
9 Financial assumptions – expected future experience – typical funding basis .......... 23
10 Financial assumptions – expected future experience – gilts basis ........................... 25
11 Financial assumptions – expected future experience – best estimate basis ............ 27
12 Derived defined benefit contracted-out rebates........................................................ 34
13 Other practical issues ............................................................................................... 36
14 Process for decision ................................................................................................. 39
15 Revaluation rate for Guaranteed Minimum Pensions ............................................... 40
Appendix A Legal background ..................................................................................................... 41
Appendix B Assumptions adopted at the previous review ........................................................... 44
Appendix C Assumptions adopted at this review ......................................................................... 45
Appendix D Evidence in support of the financial assumptions .................................................... 48
Appendix E Evidence in support of the mortality assumptions ................................................... 60
Appendix F Response to consultation ......................................................................................... 62
Review of rebates and reduced rates of National Insurance contributions
for members of contracted-out defined contribution pension schemes
1 Overview .................................................................................................................. 75
2 Introduction to contracting out .................................................................................. 77
3 Deriving the defined contribution contracted-out rebates ......................................... 79
Appendix A Legal background ..................................................................................................... 83
Appendix B Contracted-out rebates – COMPS ........................................................................... 86
Appendix C Contracted-out rebates – APPs ................................................................................ 87
Appendix D Assumptions adopted at the previous review. .......................................................... 88
Report by Secretary of State for Work and Pensions in accordance with
sections 42(1)(b), 42B(1)(b) and 45A(1)(b) of the Pension Schemes Act 1993
1 Report ...................................................................................................................... 91
Review of rebates and reduced rates of National
Insurance contributions for members of defined
benefit contracted-out pension schemes
Report to the Secretary of State for Work and
Pensions by the Government Actuary
Date: 21 January 2011
Author: Trevor Llanwarne
The Right Hon Iain Duncan Smith MP
Secretary of State for Work and Pensions
Department for Work and Pensions
Caxton House
Tothill Street
London
SW1H 9NA
21 January 2011
Dear Secretary of State
Review of rebates and reduced rates of National Insurance contributions for members
of defined benefit contracted-out pension schemes
I am pleased to provide this report setting out my review of the rebates and reduced rates of
National Insurance contributions for members of contracted-out pension schemes for the
2012 – 2017 quinquennium. This report addresses the duties in respect of contracting-out
placed upon me in my role as Government Actuary by the Pensions Act 1993.
Due to the expected abolition of contracting-out on a defined contribution basis, my report
focuses on the defined benefit contracted-out rebate. A supplementary report covering
defined contribution contracted-out rebates will be provided separately.
My report sets out defined benefit contracted-out rebates calculated using three alternative
valuation approaches. These show a range of possible outcomes for the rebate. Its intention
is to enable you to make your decision on the rebate, based on the policy you wish to
implement, on what you wish the rebates to deliver.
Yours sincerely
Trevor Llanwarne
Government Actuary
Finlaison House 15-17 Furnival Street London EC4A 1AB T: +44 (0)20 7211 2600 W: www.gad.gov.uk
Direct line: +44 (0)20 7211 2620 Fax number: +44 (0)20 7211 2601 Email address: trevor.llanwar[email protected].uk
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
1 Overview
1.1 This report is addressed to the Secretary of State for Work and Pensions and sets
out my review of the rebates and reduced rates of National Insurance contributions
for members of defined benefit contracted-out pension schemes for the 2012 – 2017
quinquennium.
1.2 Individuals may contract out through membership of:
>
a Contracted-Out Salary Related Scheme (COSRS),
a Contracted-Out Money Purchase Scheme (COMPS), or
an Appropriate Personal Pension (APP).
>
>
1.3 Those who contract-out through the first arrangement are considered to be
‘contracted-out on a defined benefit basis’. Those who contract out through the
second and third arrangements are considered to be ‘contracted-out on a defined
contribution basis’.
1.4 The Secretary of State is required by statute to lay before Parliament, at intervals of
not more than five years, a report setting out the Government Actuary’s opinion on:
>
For COSRS, changes since the last report in the factors affecting the cost of
providing benefits of equivalent actuarial value to the Additional Pension that is
forgone by workers who are contracted-out, and
>
For COMPS and APPs, the percentages required to reflect the cost of providing
benefits of equivalent actuarial value to the Additional Pension that is forgone by
workers who are contracted-out.
1.5 At the request of the Department for Work and Pensions, my report is extended to
include advice on the appropriate level of the contracted-out rebate for those
contracted-out on a defined benefit basis.
1.6 At the request of the Department for Work and Pensions, I should also propose a
rate of revaluation of Guaranteed Minimum Pensions for contracted-out workers who
leave pensionable service in the period 2012 to 2017.
1.7 My advice in this report has been informed by public consultation.
1.8 Enabling legislation now provides for the abolition of contracting-out on a defined
contribution basis. The abolition date is expected to be 6 April 2012. Therefore, this
report relates to the defined benefit contracted-out rebate, though I have also set out
the current legislative position in respect of contracting-out on a defined contribution
basis. I am required by legislation to propose defined contribution contracted-out
rebate percentages for, as a minimum, the year 2012/13, even though it is expected
that they will not come into force. These proposed rebates will be set out in a
separate report.
1
>
>
A
>
>
>
>
>
>
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
1.9 The defined benefit contracted-out rebate for 2007 to 2012 is 5.3%. This is split
1.6% employee and 3.7% employer. The rebate percentages are of earnings
between the Lower Earnings Limit and the Upper Accrual Point. At the previous
review the Government Actuary proposed a rebate of 5.8%, but the Secretary of
State decided not to accept this proposal, and instead retained the rate of 5.3%
which had been recommended and adopted for 2002 to 2007.
1.10 In my opinion, since the previous review the main relevant changes to the factors
which affect the cost of providing benefits of equivalent actuarial value to the
dditional Pension forgone by those contracted-out on a defined benefit basis are:
the profile of members of private sector COSRS is changing as a result of
continuing defined benefit scheme closures to new entrants and new accrual,
the future existence of many schemes is now likely to be curtailed which might
lead to different financial considerations than those that would apply to schemes
which do not plan to wind-up,
the operation of the Pension Protection Fund and the Employer Debt
Regulations are now more clear,
state pension age has been raised for many people, which reduces the amount
of Additional Pension forgone by contracted-out members each year, and
means that it is deducted later,
expectations of members’ longevity continue to rise,
a new funding regime for defined benefit schemes is now fully in place (deriving
from the Pensions Act 2004),
typical investment portfolios for funded pension schemes have become more
prudent,
developing actuarial opinion,
economic conditions are noticeably different, with possible implications for
expected future investment returns and inflation, and
the state Additional Pension will be indexed in future by reference to the
Consumer Prices Index, rather than by the Retail Prices Index.
1.11 In respect of my advice on the appropriate level of the defined benefit contracted-out
rebate, I offer three alternative valuation approaches, depending on the policy you
wish to adopt on what you wish the rebates to deliver.
1.12 The three alternative valuation approaches are:
a ‘best estimate’ basis, which in my view gives a number which in broad terms
is equally likely to be too much as too little in the long term for the scheme to
provide benefits at the level equivalent to the reduction in state pension,
a ‘typical funding’ basis, designed to give a number consistent with a prudent
approach to funding adopted by typical schemes, and
2
>
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
>
a ‘gilts’ basis, designed to give a number consistent with a self-sufficiency
approach.
1.13 Based on the assumptions outlined in this report, the derived rebates are:
Using state pension ages based on existing legislation:
4.9% on a ‘best estimate’ basis,
6.3% on a ‘typical funding’ basis, and
10.5% on a ‘gilts’ basis.
Using state pension ages based on expected revisions to existing legislation:
>
4.8% on a ‘best estimate ‘basis,
6.2% on a ‘typical funding’ basis, and
10.4% on a ‘gilts’ basis.
1.14 The decision to set the rebate for 2012 to 2017 rests with the Secretary of State. I
suggest that he might reach his decision by:
>
>
>
taking a view on the approach that fits the policy that it is desired to adopt for
this review (for example, any of the three bases described above, or indeed any
variant in between or outside the range offered here),
considering, if appropriate, any refinements (for example, on state pension age,
the extent (if any) to which guarantee aspects should be allowed for, margins,
rounding, etc.), and
determining the rebate and the split between employer and employee (noting
that this aspect is outside of any considerations in this report).
1.15 I have also been asked to propose a rate for the revaluation of Guaranteed Minimum
Pension (GMP) for contracted-out members who leave pensionable service in the
period 2012 to 2017. I propose that this rate is set at 4.75% pa. This compares to a
current rate of 4.0% pa.
1.16 Other than the Secretary of State for Work and Pensions, no person or third party is
entitled to place any reliance on the contents of this report. Neither I nor anyone in
GAD accepts any liability to any other person or third party for any act or omission
taken, either in whole or part, on the basis of this report.
3
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
2 Introduction to contracting out
In this section I provide a general introduction to contracting-out, for both defined benefit
and defined contribution schemes.
2.1 Since the introduction of the state earnings-related pension scheme (SERPS) in
1978, employees have been able to ‘contract-out’ – that is to give up all or part of
their state Additional Pension benefits. In return they pay lower National Insurance
contributions, or receive rebates of National Insurance contributions paid into their
defined contribution pension schemes or personal pensions.
2.2 The principle of contracting-out continued after the introduction of the state second
pension (S2P), which replaced SERPS in April 2002.
2.3 The legal background of contracting-out is set out in Appendix A. This gives the
relevant statutory references and a summary of the effect on members of
contracting-out.
The contracted-out rebate
2.4 The ‘contracted-out rebate’ refers to the effective reduction of National Insurance
contributions for members of pension schemes, and their employers, which are
contracted-out of the state Additional Pension.
2.5 The contracted-out rebate is set having regard to the cost of providing benefits of
equivalent actuarial value to the state Additional Pension that is forgone by workers
who are contracted-out.
COSRS, COMPS and APPs
2.6 Individuals may contract out through membership of:
>
a Contracted Out Salary Related Scheme (COSRS),
a Contracted Out Money Purchase Scheme (COMPS), or
an Appropriate Personal Pension (APP).
2.7 Those who contract out through the first arrangement are considered to be
‘contracted-out on a defined benefit basis’. Those who contract out through the
second and third arrangements are considered to be ‘contracted-out on a defined
contribution basis’.
2.8 In respect of members of COSRS, both the employer and employee pay reduced
rates of National Insurance contributions.
2.9 For members of APPS, HM Revenue & Customs makes payments directly to the
pension scheme in respect of contracted-out members.
4
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
2.10 A combination of the above occurs for members of COMPS. Reduced rates of
National Insurance are paid and the employer pays the balance of the contracted-
out rebates into the pension scheme. Age related top up payments are also made
to the pension scheme by HMRC.
2.11 A single contracted-out rebate percentage is applied in respect of all members of
COSRS.
2.12 Tables of age related rebates are applied in respect of members of COMPS and
APPs.
Abolition of contracting-out on a defined contribution basis
2.13 Enabling legislation now provides for the abolition of contracting-out on a defined
contribution basis. The abolition date is expected to be 6 April 2012. Therefore, this
report restricts itself to determining the issues and costs relating to the defined
benefit contracted-out rebate. I am required by legislation to propose defined
contribution contracted-out rebate percentages for, as a minimum, the year 2012/13,
even though it is expected that they will not come into force. These proposed
rebates will be set out in a supplementary report which will be provided separately.
The Government Actuary’s and the Secretary of State’s obligations
2.14 The Secretary of State is required by statute to lay before Parliament, at intervals of
not more than five years, a report setting out the Government Actuary’s opinion on:
For COSRS, changes since the last report in the factors affecting the cost of
providing benefits of equivalent actuarial value to the Additional Pension that is
forgone by workers who are contracted-out, and
For COMPs and APPs, the percentages required to reflect the cost of providing
benefits of equivalent actuarial value to the Additional Pension that is forgone by
workers who are contracted-out.
2.15 At the request of the Department for Work and Pensions, my report is extended to
include advice on the appropriate level of the contracted-out rebate for members of
COSRS.
2.16 It has become established practice for the Government Actuary’s advice to be
informed by public consultation, as discussed in sections 3.6 to 3.10.
2.17 My role, as Government Actuary, is limited to providing advice on the contracted-out
rebates. The power to decide and set the rebate rates (and the split between the
employee and employer) is vested in the Secretary of State for Work and Pensions.
2.18 In addition, I have been asked by the Department for Work and Pensions to propose
a rate for the revaluation of Guaranteed Minimum Pension (GMP) for contracted-out
members who leave pensionable service in the period 2012 to 2017. This is
covered in section 15.
5
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
3 The 2012 – 2017 defined benefit contracted-out rebate review
In this section I provide a general overview of the process I have adopted to conduct this
review of the National Insurance rebates for contracted-out staff.
Approach to this review
3.1 This report relates to the contracted-out rebate for those contracted-out on a defined
benefit basis (the ‘defined benefit contracted-out rebate’).
3.2 At previous reviews, the Government Actuary suggested a single set of assumptions
to value the benefits forgone by contracted-out workers, which led to a single
proposed value for the defined benefit contracted-out rebate.
3.3 Until the last review, in 2005/06, the Secretary of State had always accepted the
defined benefit contracted-out rebate proposed by the Government Actuary.
However, at the last review the Secretary of State instead set the rebate at a
somewhat lower level than that proposed, citing the state of the public finances.
3.4 I have adopted a different approach for this review. Following discussions with the
Department for Work and Pensions and HM Treasury and after considering the
responses to the public consultation, I am advising on three alternative valuation
approaches. These show a range of possible outcomes for the defined benefit
contracted-out rebate, so that the Secretary of State can make his decision on its
value with a better understanding of the possible approaches, outcomes and
implications.
3.5 The three alternative valuation approaches are:
>
a ‘best estimate’ basis, which in my view gives a number which in broad terms
is equally likely to be too much as too little in the long term for the scheme to
provide benefits at the level equivalent to the reduction in state pension,
a ‘typical funding’ basis, designed to give a number consistent with a prudent
approach to funding adopted by typical schemes, and
a ‘gilts’ basis, designed to give a number consistent with a self-sufficiency
approach.
Public consultation
3.6 As part of this review, the Government Actuary’s Department issued a public
consultation document that put forward possible assumptions for determining the
defined benefit contracted-out rebate on ‘best estimate’, ‘typical funding’ and ‘gilts’
bases and invited comments on the proposals. The objective was to collect
constructive evidence-based comments to help refine my proposed assumptions.
3.7 The public consultation is now closed. Twelve responses were received, which I
have considered carefully before producing this report.
6
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
3.8 The responses were generally constructive and largely supportive of my proposed
approach. Detailed comments were offered on appropriate assumptions. I am
grateful to all those who contributed responses to this consultation.
3.9 A summary of the key points from the consultation replies can be found in Appendix
F.
3.10 The full public consultation document can be accessed on GAD’s website.
7
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
4 Factors affecting the cost of providing contracted-out benefits
In this section I comment on the changes to the factors which affect the cost of providing
benefits of equivalent actuarial value to the Additional Pension forgone by contracted-out
workers. This addresses my statutory obligations in respect of defined benefit schemes.
4.1 In respect of the defined benefit contracted-out rebate, my statutory obligation is to
comment on the changes to the factors which affect the cost of providing benefits of
equivalent actuarial value to the Additional Pension forgone by contracted-out
workers. In my opinion, the main relevant changes since the previous review are:
>
>
>
>
>
>
>
>
>
>
the profile of members of private sector COSRS is changing as a result of
continuing defined benefit scheme closures to new entrants and new accrual,
the future existence of many schemes is now likely to be curtailed which might
lead to different financial considerations than those that would apply to schemes
which do not plan to wind-up,
the operation of the Pension Protection Fund and the Employer Debt
Regulations are now more clear,
state pension age has been raised for many people, which reduces the amount
of Additional Pension forgone by contracted-out members each year, and
means that it is deducted later,
expectations of members’ longevity continue to rise,
a new funding regime for defined benefit schemes is now fully in place (deriving
from the Pensions Act 2004),
typical investment portfolios for funded pension schemes have become more
prudent,
developing actuarial opinion,
economic conditions are noticeably different, with possible implications for
expected future investment returns and inflation, and
Additional Pension will be indexed in future by reference to the Consumer
Prices Index, rather than by the Retail Prices Index.
4.2 The Pensions Act 2007 and the Pensions Act 2008 contain further reforms to the
state second pension. These result in a number of technical changes to the
calculation of the state second pension and contracted-out benefits. These do not
have a material effect on the defined benefit contracted-out rebate percentage.
4.3 The factors set out in paragraph 4.1 and 4.2 are discussed further below.
8
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Member profile
4.4 The relevant population of contracted-out workers includes both public and private
sector pension schemes. There is a trend in the private sector towards closing
defined benefit schemes to new entrants. Those which remain open to accrual by
existing members therefore experience a gradual increase in the average age of the
members.
4.5 Since the value of pensions to workers in service generally increases with age, then
this gradual increase in the average age of workers in the population leads to a
gradual increase in the derived rebate.
4.6 The illustrated defined benefit contracted-out rebate values discussed in section 12
are based on a representative sample of contracted-out workers, taken from an
extract from National Insurance records.
Anticipated curtailment of defined benefit schemes
4.7 There is evidence highlighted by the consultation responses that many sponsors of
private sector defined benefit pension schemes are considering closing their
schemes to all future accrual (including that for existing members). It follows that
some of these schemes may then look to buy-out their benefits with an insurance
company.
4.8 Pension schemes targeting buy-out tend to hold less risky assets which more
closely match their liabilities (gilts and corporate bonds) and hence fund for their
benefits on a more prudent basis than ongoing schemes. In turn, this may cause
the perceived costs of providing benefits to rise.
Employer debt regulations and the Pension Protection Fund
4.9 At the previous review the Government Actuary at the time identified the introduction
of the Pension Protection Fund (PPF) and revisions to the employer debt regulations
as factors affecting the cost of providing benefits contracted-out on a defined benefit
basis. However, the implications of these were not known at the time.
4.10 These factors could be considered to have changed the nature of the ‘pension
promise’, increasing the extent to which it could be considered ‘guaranteed’ outside
of the pension scheme and therefore potentially reducing its direct perceived cost to
the sponsoring employer (or, arguably, the opposite view might be taken that this
increases the perceived cost). The extent to which it is appropriate to consider the
value of this guarantee in the defined benefit contracted-out rebate is discussed
further from paragraph 13.8.
Change of state pension age
4.11 Since the last review, state pension age for younger workers has been increased
from 65, ultimately to 68 for the youngest workers. This has two effects:
Because the state pension (and hence the state pension forgone by contracted-
out workers) is payable later, then it is less valuable in most circumstances.
This is because the payments are further into the future and also payable for a
shorter period, and
>
9
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
>
Because of the design of the contracted-out deduction
1
, a later state pension
age flows through to a lower rate of accrual.
4.12 Both these features generally lead to a reduction in the appropriate rebate.
4.13 As part of the 2010 spending review, the Government announced plans to
accelerate the rise in state pension ages. Subject to Parliamentary approval the
state pension age should rise to age 65 by 2018 and age 66 by 2020.
4.14 I have been asked by the Department for Work and Pensions to provide two sets of
rates for the defined benefit contracted-out rebate – one set of rates based on
existing legislation and the other on the assumption that the changes to state
pension ages will receive Parliamentary approval.
4.15 Further rises to the state pension age beyond those described in paragraph 4.13
above would make the rebate more generous, all other things being equal.
Expected longevity
4.16 Longevity has been improving for several decades, though we cannot be certain that
this trend will continue in future. Nevertheless, demographers and actuaries
generally expect greater improvements in future longevity than they did previously.
4.17 Allowance for greater improvements in future longevity than previously anticipated
leads to an increase in derived rebates.
New funding regime
4.18 The funding requirements of the Pensions Act 2004 came into force shortly before
the previous review. While their intent was then known, their impact was unclear.
4.19 The Act provided for a new approach to pension scheme funding. It also introduced
the Pensions Regulator, whose code of practice and guidance on funding has
subsequently had a material impact on the pension industry’s thinking and practice.
4.20 The increased emphasis on prudence in funding practice, as required by the
Pensions Act 2004, may have had an effect on pension scheme investment
strategy, valuation assumptions, contribution rates and technical provisions.
4.21 Arguably, changes in funding practice do not affect the underlying cost of the
benefits being provided. However, to the extent that changes in practice do appear
to have led to greater advance funding of pension benefits, or more prudence in
typical funding, then the perceived cost of providing benefits will have risen in some
circumstances.
Typical investment portfolios
4.22 Typical investment portfolios for funded pension schemes have become more
prudent since the previous review.
1
The contracted-out deduction is the reduction in Additional Pension applied to individuals who have
accrued pension in a contracted-out arrangement.
10
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
4.23 The Purple Book 2010 (published by the PPF and the Pensions Regulator) shows
(in table 7.1 on page 67) that for PPF eligible defined benefit schemes the average
equity allocation has fallen from 61.1% in 2006 to 42.0% in 2010. The average gilts
and fixed interest portion has increased from 28.3% in 2006 to 40.4% in 2010.
4.24 To the extent that more prudent investment strategies are allowed for in the
assumptions that are used to derive the defined benefit contracted-out rebate, the
rebate would be higher.
Actuarial opinion
4.25 Actuarial opinion continues to develop. This is partly due to the introduction of the
new funding regime, and partly due to separate developments in the actuarial field.
The tendency is for the costs of pension provision in terms of typical funding to be
put onto a more prudent footing (that is to be valued more highly).
Current economic conditions
4.26 Economic conditions are materially different now compared with the previous review.
The long-term impact of recent events (the ‘credit crunch’ and other significant
problems in the banking system) is not yet clear. It seems that expectations of
future long-term investment returns may now be lower than previously anticipated.
This would increase the derived rebates.
Choice of index for increases in payment
4.27 At the June 2010 budget, the Chancellor of the Exchequer announced that in future
the state second pension (among other things) would increase in payment in line
with the Consumer Prices Index, rather than the Retail Prices Index. In general, this
is expected to lead to smaller pension increases, which reduces the value of the
state second pension, and hence leads to smaller derived rebates.
Minor changes in Pensions Act 2007 and Pensions Act 2008
4.28 The technical changes to the calculation of the state second pension are minor. For
a few individuals, the deduction from state benefits as a consequence of contracting-
out will exceed the state pension that they would otherwise receive, so that their
actual state second pensions are reduced to zero. This does not have a material
effect on the calculation of the defined benefit rebate.
11
>
>
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
5 Deriving the defined benefit contracted-out rebate
In this section I explain how a rebate rate is derived. The derived rebate rate depends on:
>
the State benefits forgone by contracted-out staff,
the assumptions adopted to value these benefits, and
the method adopted to perform the valuation calculation.
State benefits forgone by contracted-out staff
5.1 To derive a defined benefit contracted-out rebate we have to identify the benefits in
the state second pension scheme which are forgone by contracted-out staff.
Effectively, these are calculated in the same way as previously under the State
Earnings Related Pension Scheme (SERPS). That is:
>
a percentage of earnings in a defined band,
payable from state pension age,
revalued in deferment in line with average earnings growth,
increased in payment in line with prices inflation, and
with contingent spouse’s benefits of up to 50% of the contributor’s pension.
5.2 Specimen benefit rates are shown in the table below:
Year beginning Age at
State pension Annual accrual
beginning of
age of benefits *
year
6 April 2012 30 68 0.385%
40
50
67
66
0.392%
0.408%
60
60
62 (women)
65 (men)
0.588%
0.526%
6 April 2016 30
40
50
68
67
66
0.385%
0.392%
0.400%
60 65 0.476%
* based on salary between the Lower Earnings Limit the Upper Accrual Point
12
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Valuation assumptions
5.3 In order to place a value on the state benefits forgone by contracted-out staff, we
must adopt assumptions in respect of expected future demographic and financial
experience. The demographic assumptions include:
>
>
>
>
>
mortality for members of contracted-out pension arrangements, both before and
after state pension age (noting that longevity after state pension age is much
more significant than mortality before then),
mortality for surviving spouses of members of contracted-out pension
arrangements,
proportions of members of contracted-out pension arrangements who are either
married or have a civil partner at death,
probability of ‘remarriage’ for surviving spouses (since inherited benefits cease
on remarriage), and
age differences between members of contracted-out pension arrangements and
their partners.
5.4 The financial assumptions are used to derive discount rates which are used to
derive capitalised values of the expected future state benefit payments.
5.5 All the assumptions are applied so as to recognise that they apply for the period
2012 to 2017.
5.6 There are a range of assumptions which might be adopted to value benefits,
depending on the purpose of the valuation. For this report, I have illustrated the
rebates which would be derived from a range of assumption sets which reflect
alternative approaches to valuing the state benefits forgone. These alternative
valuation approaches are discussed in section 6.
5.7 The derivation of demographic assumptions is discussed in section 7. The
derivation of discount rates is discussed in sections 8 to 11.
Calculation methodology
5.8 I have derived assumptions which can be applied to value the state benefits forgone
as a result of contracting out for a given individual. These state benefits vary from
individual to individual, based on their sex and year of birth.
5.9 There is no need to make any assumptions about individuals’ earnings. This is
because the state benefits forgone and the National Insurance rebates are all
quoted as percentages of earnings between the Lower Earnings Limit and the Upper
Accrual Point. The rebate is thus the same percentage of these earnings,
regardless of the size of the earnings.
5.10 I have therefore calculated the theoretical rebate on each of my assumption sets for
men and women at each age up to state pension age. These rebates are then
aggregated using weightings based on the age profile of the contracted-out
workforce. This produces a single rebate value for the workforce as a whole.
13
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
6 Valuation approaches / assumption sets
In this section I explain the three separate approaches to valuing the State benefits
forgone by contracted-out staff. I describe the approaches (which each reflect different
sets of valuation assumptions) and comment on the implications of each.
I also discuss some alternative possible valuation approaches which I have chosen not to
use in this report.
6.1 While there is a range of actuarial assumptions which may be used when valuing
pension benefits, I propose to offer three alternative valuation approaches designed
to illustrate a range of potential rebate values. These are:
>
a ‘best estimate’ basis, which in my view gives a number which in broad terms
is equally likely to be too much as too little in the long term for the scheme to
provide benefits at the level equivalent to the reduction in state pension,
a ‘typical funding’ basis, designed to give a number consistent with a prudent
approach to funding adopted by typical schemes, and
a ‘gilts’ basis, designed to give a number consistent with a self-sufficiency
approach.
6.2 In my view, any of these three approaches can meet the description appearing in
legislation of equivalent actuarial value (see paragraph 1.4), but any rebate at a level
below that derived from the best estimate approach should not be considered as
such. This matter is analysed further in section 11.
6.3 The three bases can be described and rationalised as follows.
Best estimate basis
6.4 A ‘best estimate’ basis is intended to lead to a rebate which is considered, in broad
terms, equally likely to deliver benefits which are more or less than the state second
pension forgone as a consequence of being contracted-out. A rebate on a ‘best
estimate’ basis uses estimates of the relevant assumptions which do not include any
material margins of prudence and which are applied in a way which reflects the
actual investment strategies adopted by such schemes. A rebate based on a ‘best
estimate’ basis is expected to be sufficient, half the time, to cover the cost of
providing benefits equivalent to the state second pension forgone by contracted-out
workers.
14
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Typical funding basis
6.5 A ‘typical funding’ basis should include margins of prudence (relative to the ‘best
estimate’ basis) which are consistent with the regulatory regime imposed by the
Pensions Act 2004 and the Pensions Regulator’s guidance. Thus, the rebate on the
‘typical funding’ basis represents the amount which funded defined benefit schemes
would typically hold in practice in order to cover benefits of equivalent actuarial value
to those forgone as a result of contracting-out (where the actuarial value is
measured in the same way as technical provisions are calculated). A rebate based
on ‘typical funding’ assumptions is expected to be more likely than not to be
sufficient to cover the long-term cost of providing benefits equivalent to the state
second pension forgone by contracted-out workers.
Gilts basis
6.6 The ‘gilts’ basis is intended to lead to a rebate which would allow a contracted-out
defined benefit pension scheme to provide benefits equal to those forgone with a
high degree of certainty, by investing in gilts. It does not mirror the cost of
purchasing annuities from an insurance company, nor include any additional
reserves of capital to meet extreme adverse outcomes. Thus, the rebate on the
‘gilts’ basis represents the amount which funded defined benefit schemes could hold
in order to cover benefits of equivalent actuarial value to those forgone as a result of
contracting-out, if they adopted a relatively low risk investment strategy based on
gilts. Some describe this approach as ‘self-sufficiency’, since it may be possible at
this level to avoid the need for extra funding in adverse times. A rebate reflecting
the ‘gilts’ basis is expected to be sufficient on average to allow a scheme to adopt a
low risk investment strategy and still cover the forgone state benefits without the
sponsor having to provide additional financial support.
Values for the assumptions under each basis
6.7 Each basis consists of a set of assumptions that reflect the characteristics of the
basis. The assumptions used for this review are summarised in Appendix C. (For
comparison, the assumptions used to calculate the recommended defined benefit
contracted-out rebate at the last review are included as Appendix B). The derivation
of the financial and demographic assumptions is set out in sections 7 to 11 with
supporting evidence in Appendix D and Appendix E. The responses to public
consultation (which were substantially around assumptions and approach) are
summarised in Appendix F.
Other possible bases (which have not been adopted)
6.8 A prudent basis might replicate the assumptions that insurance companies use to
price annuities. However, these are not published. Buy-out terms are also
influenced by market forces as well as the expected cost of providing benefits, and
have been subject to considerable volatility in recent years. Hence I have not
attempted to produce figures on a buy-out basis.
15
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
6.9 Four respondents to the consultation supported the inclusion of an ‘accounting’
basis in our final report (although two respondents explicitly rejected an ‘accounting’
basis). Given that accounting assumptions are intended to represent an overall best
estimate basis, I have decided against the inclusion of yet another basis which might
well be a duplicate.
6.10 A ‘SCAPE’
2
basis was also suggested by one consultation respondent. However,
because I am focusing on the ‘cost’ in the private sector of providing the state
second pension benefits forgone, and because public sector contracting out does
not impact at the employer level on the PSBR (Public Sector Borrowing
Requirement), tax, or real cash costs to the economy, a SCAPE basis would have
little relevance.
2
SCAPE is short for Superannuation Contributions Adjusted for Past Experience, and is the
methodology used to set ‘inter-Government’ budget charges for the unfunded public sector pension
schemes.
16
>
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
7 Demographic assumptions
In this section I discuss the general approach to the demographic assumptions which I
have adopted. There is considerable common ground between the assumptions adopted
for each of my three valuation approaches, so all are discussed together.
Assumptions for administration expenses are also discussed in this section.
7.1 Five specific sets of demographic assumptions are potentially needed in order to
calculate the defined benefit contracted-out rebate:
mortality for members of contracted-out pension arrangements, both before and
after state pension age (noting that longevity after state pension age is much
more significant than mortality before then),
mortality for surviving spouses of members of contracted-out pension
arrangements,
proportions of members of contracted-out pension arrangements who are either
married or have a civil partner at death,
probability of ‘remarriage’ for surviving spouses (since inherited benefits cease
on remarriage), and
age differences between members of contracted-out pension arrangements and
their partners.
7.2 State pension arrangements (as they affect future accrual) make little distinction
between the benefits payable to widows, widowers and surviving civil partners.
‘Spouse’, ‘marriage’ and ‘remarriage’ should be construed accordingly in the
following discussion.
7.3 I consider that ‘best estimate’ assumptions for proportions married and spouses’ age
differences can also be used in the ‘typical funding’ basis and ‘gilts’ basis without
causing difficulties of validity or interpretation. These are not particularly sensitive
assumptions and adequate margins of prudence may be reflected in the mortality
and financial assumptions.
7.4 I consider it appropriate that the mortality assumptions should distinguish between a
‘best estimate’ basis and ‘typical funding’ basis. I do not consider that using different
mortality assumptions for the ‘gilts’ basis to the ‘typical funding’ basis would provide
additional information.
Mortality
7.5 The Government Actuary’s Department has derived ‘best estimate’ mortality
assumptions based on the experience of the contracted-out workforce by looking at
experience from 2000/01 to 2007/08 and comparing against standard tables. I am
happy that this analysis gives me a ‘best estimate’ fit as follows:
17
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Base table
3
Male mortality
S1PML
Female mortality
S1PFL
Mortality of spouses of men
S1DFL
Mortality of spouses of women
S1PML
Note that there is no table S1DML produced for widowers.
7.6 Given the limited data available on civil partners, I have used the same spouse’s
mortality assumptions (as above) for bereaved civil partners.
7.7 I have allowed for future improvements in mortality in line with the 2008-based UK
principal population projections as produced by the Office for National Statistics
demography unit. The standard mortality tables set out in paragraph 7.5 above
were adjusted to 2008 from their base year in line with improvements in UK
population mortality. While the Office for National Statistics does not express an
opinion on the likelihood that this projection will be borne out in practice, I believe
that it can reasonably be adopted as a best-estimate assumption consistent with
many other situations where this approach is adopted.
7.8 For the ‘typical funding’ basis I have incorporated a two year age deduction from the
base mortality tables, to provide a margin of prudence.
7.9 I have also adopted a two year offset for the ‘gilts’ basis.
Proportions married (including civil partnerships)
7.10 I have used proportions married assumptions based on the 2006-based projections
as prepared by the Office for National Statistics for England and Wales. I consider
these projections to be ‘best estimate’, in the absence of any better data to the
contrary. Sample proportions married at death for 2014 are given in the following
table:
Age Men Women
60
70% 67%
65
73% 67%
70
75% 62%
75
73% 52%
80
68% 39%
3
The base tables are available on the Institute and Faculty of Actuaries’ website.
18
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Probability of remarriage
7.11 At previous reviews, assumptions were made in regard to the remarriage of
surviving spouses (including civil partners). If the same assumptions were applied at
this review, it would reduce the value of the benefits by less than 1%, which I
consider is not material in this context. I have therefore omitted this assumption
from all of my bases.
Age differences
7.12 I have assumed that spousal age differences are broadly in line with the experience
of the population of Great Britain as revealed by the 2001 census, which is the latest
available. Sample age differences are set out in the following table:
Age Men Women
60
spouse 3 years younger spouse 2 years older
65
spouse 3 years younger spouse 2 years older
70
spouse 4 years younger spouse 2 years older
75
spouse 4 years younger spouse 1 year older
80
spouse 4 years younger spouse same age
7.13 At first glance it might seem surprising that the age differences for men and women
are not equal and opposite, but this a natural combination of the age differences
within marriages and higher male mortality rates, which lead to this result.
7.14 The 2001 census data concerns opposite sex relationships only, given its date. Of
course, civil partners are on average the same age as each other (though the elder
is more likely to die first on average).
7.15 I consider that the use of average spousal age differences based on the 2001
census remains satisfactory, with the passage of time and inclusion of a small
proportion of civil partners and workers in Northern Ireland not causing a material
issue.
Weights
7.16 The reduction in the National Insurance contributions for individuals contracted-out
on a defined benefit basis will continue to be a single percentage of band earnings,
independent of sex and age. It is necessary, therefore, to weight the derived
individual age and sex related defined benefit contracted-out rebates to obtain an
appropriate average rebate.
7.17 The weightings needed are to ascertain what percentage of the total population will
fall into. Five-year age bands for each sex have therefore been derived.
19
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
7.18 The weightings are based on sample data from the National Insurance Fund using
trends in data projected over the period the new defined benefit contracted-out
rebate will apply.
7.19 In the absence of any data specific to private sector schemes, I have been obliged
to use data based on members of both private and public sector schemes. This is
discussed further from paragraph 13.14.
7.20 The weights used at this review are given in the following table:
Age Group Men Women
20-24
0.83% 1.09%
25-29
3.49% 5.03%
30-34
4.88% 5.88%
35-39
5.60% 6.04%
40-44
7.05% 7.71%
45-49
8.57% 9.89%
50-54
8.32% 9.37%
55-59
5.95% 6.09%
60-64
2.59% 1.62%
Total 47.28% 52.72%
Expenses – administrative
7.21 For defined benefit schemes, I am concerned with the marginal cost of contracting-
out. The National Insurance rebate should allow for any specific expenses involved.
There are some minor expenses incurred solely as a result of contracting-out which
should be loaded onto the rebate.
7.22 In the absence of any evidence to the contrary, I have retained the expense
allowance of 0.2% of band earnings made at the last review, for each of my bases.
This assumption was considered reasonable by respondents to the consultation.
Evidence
7.23 Evidence in respect of the mortality assumptions is contained in Appendix E.
Evidence provided in the response to the public consultation is discussed in
Appendix F.
20
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
8 Financial assumptions – asset allocation
My proposed ‘best estimate’ basis requires assumptions about the average asset
allocation decisions made by typical defined benefit pension schemes. In this section I
discuss this preliminary issue, prior to deriving usable discount rates (in the following
sections).
8.1 The ‘best estimate’ basis requires assumptions about the average asset allocation
effectively adopted by typical funded defined benefit pension schemes for members
of different ages. In practice, many schemes base their asset allocation on the
overall mix of members. This section therefore explains how we have interpreted
the available evidence in order to derive a typical portfolio relative to differing age
profiles so that we can then use the age weightings derived in section 7 to set
overall investment returns and discount the required assumptions.
8.2 The 2009 Purple Book published by the Pensions Regulator covers almost all
pension schemes in the PPF eligible UK defined benefit scheme universe. Section
7.6 (Chart 7.4) of the Purple Book
4
shows that, broadly speaking:
schemes with around 90% non pensioner and 10% pensioner liabilities have an
average asset allocation of approximately 65% equities and property, 30% gilts
and fixed interest and 5% cash and other,
schemes with around 50% non pensioner and 50% pensioner liabilities have an
average asset allocation of approximately 50% equities and property, 40% gilts
and fixed interest and 10% cash and other, and
schemes with around 10% non pensioner and 90% pensioner liabilities have an
average asset allocation of approximately 20% equities and property, 70% gilts
and fixed interest and 10% cash and other.
8.3 Whilst not uniformly consistent, these figures are broadly consistent with an asset
allocation of between 80%/20% and 90%/10% equities/bonds for non-pensioner
liabilities and 100% bonds for pensioner liabilities.
8.4 This does not provide evidence regarding differential investment strategies within
the non-pensioner group. However, anecdotally it is common to adopt a more
conservative investment strategy the closer the active members are to retirement.
This is also consistent with the approach adopted within defined contribution
schemes where it is common to switch gradually over the immediate period (often
10 years) before retirement from an investment strategy of 100% equities 10 years
or more before retirement to a strategy of 100% bonds at retirement. Assuming
such a notional strategy for defined benefit schemes would give an overall asset
allocation consistent with the evidence above. This is therefore the assumption I
have made.
4
The recent 2010 Purple Book does not contain the equivalent evidence in a usable form.
21
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
8.5 We have further assumed a split of the bond portion of the assets of 30% index-
linked gilts, 30% conventional gilts and 40% corporate bonds. This is based on the
evidence in Section 7.3 (Table 7.3) of the 2009 Purple Book which shows an
average split of 32.6% index-linked gilts, 29.0% conventional gilts and 38.3%
corporate bonds.
8.6 The resulting average asset allocations over the invested lifetime of individuals of
different ages are given below. The consultation responses were supportive of
these assumed asset allocations.
Pre-retirement asset allocation
(A differing asset allocation has been assumed dependent on term to SPA)
Just before SPA 60% gilts (30% index linked, 30% fixed interest), 40% corporate
bonds
10 years to SPA Fully invested in equities 10 years from state pension age,
switching to 60% gilts (30% index linked, 30% fixed interest), 40%
corporate bonds over the remaining 10 years to state pension age.
On average this translates approximately to 50% equities, 30% gilts
and 20% corporate bonds over a 10 year term to state pension
age.
Over 25 years to SPA Fully invested in equities 25 years from state pension age,
switching from this to 60% gilts (30% index linked, 30% fixed
interest), 40% corporate bonds over the last 10 years to state
pension age.
On average this translates approximately to 80% equities, 12% gilts
and 8% corporate bonds over a 25 year term to state pension age.
Post-retirement asset allocation
60% gilts (30% index linked and 30% fixed interest) and 40%
corporate bonds.
8.7 More details on the evidence to support the analysis in this section can be found in
Appendix D.
22
>
>
>
>
>
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
9 Financial assumptions – expected future experience – typical funding
basis
In this section I discuss assumptions relating to future financial experience and future
financial conditions, and combine these assumptions to derive the discount rate
assumptions which are needed to derive the contracted-out rebate rates.
This section looks at the typical funding basis, and provides assumptions which lead to a
rebate of 6.3% (or 6.2% if the new State Pension Age provisions are allowed for). The
next two sections consider the gilts basis and best estimate basis.
9.1 The key financial assumptions that are required are discount rates based on:
the expected return on appropriate investments after state pension age (SPA) in
excess of annual increases on the state benefits forgone,
the expected return on appropriate investments before state pension age in
excess of the revaluation of earnings factors on state benefits forgone (by
reference to national average earnings growth), and
the expected nominal return on appropriate investments before state pension
age (used for the year immediately prior to retirement since there is no earnings
revaluation of the state benefits forgone in that year).
9.2 Since the nominal rate (third bullet point above) is only used for the year
immediately prior to retirement, it is the net rates (first and second bullet points
above) which are most important in determining the cost of provision. The nominal
discount rates are set based on assumed asset allocations.
9.3 When formulating the financial assumptions I have considered the time frame over
which they will be applied, market conditions and expectations of long run
equilibriums.
9.4 I have derived the suggested financial assumptions using a ‘building block’ approach
based on:
Consumer Prices Index (CPI),
National Average Earnings (NAE),
Gilt returns,
Corporate bond returns,
Equity returns, and
Investment expenses.
(Though not all these components are relevant for all three bases.)
23
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
9.5 The building blocks are based on market conditions as at 31 December 2010 and
have been rounded to 0.05%. I have looked at the impact of market conditions on
the derived defined benefit contracted-out rebates at each month end since
January 2010.
9.6 The figures used for the building blocks are set out below and described in the rest
of this section. Investment expenses are implicit in the quoted discount rates.
CPI (multi-year) 3.00% pa
RPI (CPI + 0.75%) 3.75% pa
NAE (RPI + 1.50%) 5.25% pa
Fixed interest gilt returns 4.65% pa
Net Pre-Retirement Discount Rate 1.19%
Net Post-Retirement Discount Rate 1.94%
Excess over gilt return to derive 1.85% (pre-retirement)
overall discount rate
0.35% (post-retirement)
Net Post-Retirement Discount Rate
9.7 In determining the ‘typical funding’ discount rate for the periods before and after
state pension age, a margin above the ‘typical funding’ fixed interest gilt return has
been assumed. This margin aims to be reflective of the average margin employed
by defined benefit pension schemes in setting their nominal discount rates for
technical provisions. I have assumed that this margin is 0.35% pa post-retirement
and 1.85% pa pre-retirement, based on the average derived from the figures shown
in the Pensions Regulator’s report set out in D.3.
9.8 For the ‘typical funding’ basis, allowing for a margin of 0.35% above gilt return
results in a nominal post-retirement return equal to 5.00% pa.
9.9 Allowing for the ‘typical funding’ CPI assumption of 3.00% pa, gives a net
post-retirement discount rate of 1.94% pa.
9.10 Investment expenses are implicit in the above ‘typical funding’ nominal rate.
Pre-Retirement
9.11 For the ‘typical funding’ basis, allowing for a margin of 1.85% pa above gilts, results
in a nominal pre-retirement return equal to 6.50% pa.
9.12 Allowing for ‘typical funding’ NAE increases of 5.25% pa results in a proposed net
‘typical funding’ pre-retirement discount rate of 1.19% pa.
Evidence
9.13 See Appendix D for more detail.
24
>
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
10 Financial assumptions – expected future experience – gilts basis
In this section I discuss assumptions relating to future financial experience and future
financial conditions in relation to the gilts basis. I derive assumptions which lead to a
rebate of 10.5% (or 10.4% if the new State Pension Age provisions are allowed for).
10.1 I have derived the suggested financial assumptions using a ‘building block’ approach
in the same way as for the typical funding basis:
>
Consumer Prices Index (CPI),
National Average Earnings (NAE),
Gilt returns,
Corporate bond returns,
Equity returns, and
Investment expenses.
10.2 The building blocks are based on market conditions as at 31 December 2010. I
have looked at the impact of market conditions on the derived defined benefit
contracted-out rebates at each month end since January 2010.
10.3 The building blocks have been rounded to 0.05% pa.
10.4 The figures used for the above building blocks are set out below and described in
the rest of this section.
Fixed interest gilt returns 4.65% pa
Excess over gilt return to derive overall Nil
discount rate
Net Pre-Retirement Discount -0.9%
Net Post-Retirement Discount Rate 1.26%
Investment expenses 0.10% pa
Net Post-Retirement Discount Rate
10.5 For the ‘gilts’ basis, the return on appropriate investments after state pension age is
determined by reference to the gilt yields, after allowing for investment expenses of
0.1% pa. This results in a nominal return equal to 4.55% pa.
10.6 Allowing for the ‘gilts’ basis CPI assumption of 3.25% pa (based on gilts break-even
inflation) gives a net post-retirement discount rate of 1.26% pa.
25
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Pre-Retirement
10.7 For the ‘gilts’ basis, the return on appropriate investments before state pension age
is 4.65% pa (the assumed fixed interest gilt yield). Deducting investment expenses
of 0.1% pa, and allowing for ‘gilts’ NAE increases of 5.50% pa, results in a proposed
net ‘gilts’ basis pre-retirement discount rate of -0.90% pa.
Evidence
10.8 See Appendix D for more detail.
26
>
>
>
>
>
>
>
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
11 Financial assumptions – expected future experience – best estimate
basis
In this section I discuss assumptions relating to future financial experience and future
financial conditions in relation to best estimate. I then consider how all the assumptions for
best estimate can be put together to derive a best estimate rebate of 4.9% (or 4.8% if the
new State Pension Age provisions are allowed for).
11.1 I have derived the suggested financial assumptions using a “building block”
approach in the same way as in sections 9 and 10 for typical funding and gilts:
Consumer Prices Index (CPI)
National Average Earnings (NAE)
Gilt returns
Corporate bond returns
Equity returns, and
Investment expenses
11.2
In order to derive an appropriate assumption set for best estimate, I am performing
an exercise which has not been attempted in previous reviews. Previously, it has
been possible to allow for broad brush margins of an overall nature such that
precision on selecting individual assumptions was not necessary. That lack of
precision cannot fit easily with best estimate.
11.3 For many assumptions, principally demographic assumptions, there is little extra
difficulty. But there are a few assumptions where there are legitimate grounds for
considering a range of possible answers given the technical analysis, the evidence
and the consultation responses. These are:
the assumption for CPI,
allowance for the fact that pensions cannot fall in times of negative inflation,
the date on which to derive yields (and other market statistics) as being
appropriate for future years,
the age profile difference between private sector contracted out workers and
public sector (see paragraphs 4.6 and 13.14), and
the remarriage assumption (see paragraph 7.11).
11.4 I now look at the issues around each of these in turn and then how to look at these
together in an overall manner.
27
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Consumer Prices Index
11.5 The issue: it is fully accepted by me (and the consultation responses did not
disagree) that it is right to assume that the Bank of England will deliver to their 2%
target on a year-by-year basis recognising that some years will be higher and some
lower. But the question I have to address is the technical one of what does 2%
year-on-year actually imply for the longer term.
11.6 For example, does 2% mean that in any one year the rate is just as likely to be
above 2% as below 2% (in which case 2% is a “median” rate and the chances are
that this implies a long-term accumulated rate equivalent to somewhat more than
2%). Or does it mean that if all years are looked at the most common rate to appear
is 2% (in which case 2% is a “mode” rate and a similar argument applies as for
median rate that the long-term equivalent is likely to be above 2%). Or if the target is
actually for the rate to average at 2% over the long-term (the “mean” rate”), then the
long-term accumulated rate should then be 2% but the chances would then be that
the Bank of England have to adjust future targeting dependent on the past average
experience so that the future may not necessarily be 2%.
11.7 The extreme position is that all such rates are the same. So at this extreme, the right
long-term rate to use would be 2%. Away from the extreme, the right approach is to
consider 2% as a minimum with no downside but with a risk of upside because of
the way the mathematics will work
5
.
11.8 The evidence:
the market implied rate for future inflation is considerably in excess of
2% – see Appendix D. The Bank of England produces regular forecasts for future
CPI over a three year period, based around the mode of the distribution. This is part
of the process by which the 2% rate is targeted. They also estimate the mean and
median of this short-term distribution. However, the Bank does not publish any
longer-term distributions or any views on how the median and mean of the longer-
term distribution would relate to the mode, nor have they expressed any view to me
over what they expect the future median and mean to be. All responses received in
our consultation, where a view has been expressed on the matter, are clear that the
rate equivalent to a Bank of England 2% annual target has to be higher than 2% for
the mean rate.
5
When considering the cumulative rate over the future long-term, we need to consider the extent to
which the distribution of future inflation rates is skewed, rather than symmetrical, and if the distribution
is considered symmetrical whether it is symmetrical on a linear basis or a different basis (for example,
a logarithmic scale). I understand that a number of people hold the view that “shocks to the system”
can cause more significant increases to the rate of inflation than decreases, without prejudicing the
annual “best estimate” of 2% year-on-year. For example, suppose in year 1 that the rate is 2%; in year
2 it doubles; in year 3 it reverts to 2%; in year 4 it halves and in year 5 it reverts again to 2%. This
would be consistent with a 2% year-on-year assumption, the median and mode are both 2% but the
cumulative rate over 5 years is 11.5% which is equivalent to an annual long-term (mean) rate of 2.2%.
And it is the mean rate that we need for our assumption.
28
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
11.9 Even so, there still remains insufficient evidence yet to support a view that the mean
should be different to the mode and the Treasury has expressed this view to us. If
we look for empirical evidence, then since the Bank of England started targeting 2%
CPI in 2003, the mean accumulated rate over the period is 2.5% compared to a
median of 2.4%. If the frequency of letters written by the Governor of the Bank of
England
6
is examined, there have been many letters written on the upside compared
to downside letters – once again suggesting that the mean will be higher than 2.0%
over a long-term.
11.10 The range: the above analysis and evidence makes it clear that the assumption I
should use has a minimum of 2.0% pa but it could be higher. If the right answer
were to be higher then I have looked at how this would change the resulting rebate
of 4.9%. At a 2.5% CPI assumption, which I believe to be the top end of the range,
there would be an addition to the rebate of up to 0.3%.
Benefits cannot have reductions when inflation goes negative
11.11 The issue: if CPI falls below zero, pensions do not suffer negative increases. This
means that the assumption for pension increases should be somewhat higher than
the assumption for CPI. If CPI has a mean value higher than the 2% year-on-year
target, then this issue is likely to be non-material. But if it is right to assume a
symmetrical distribution (see footnote on the previous page) with a mean and
median of 2%, then this issue has a bigger impact.
11.12 The range: indications suggest that this has an impact of between 0% and 0.1% as
an addition to any resulting rebate depending on what the underlying CPI
assumption is.
The date on which to derive yields
11.13 The issue: I have to make an assumption as to what the yields will be for
investments made in the period from 2012 to 2017. This is achieved by making an
assumption for the gilt return and the equity return for new investment in that period
and that depends on market prices over that period. There are a number of ways to
look at gilt returns including:
a survey of the views of respected experts, for example, economists,
use of latest available current market data on the grounds that current rates are
the latest aggregate views of buyers and sellers and if it were appropriate to
assume a shift up or down, then the price would have adjusted accordingly, and
use of a reasonably recent average yield on the grounds that the market view
has not changed but arbitrary blips are removed. Depending on the averaging
period, this does involve a view that the current price is wrong and it may well
be that this means taking a view that the market view has got it wrong rather
than just a view that there is a blip.
11.14 It is a not uncommon position of actuaries to take the middle approach on the
grounds that this is consistent with other market aspects of the valuation process.
6
The Governor has to write letters when the rate is above 3% (upside) or below 1% (downside).
29
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
11.15 The evidence: the majority of the consultation responses supported the adoption of
the middle approach.
11.16 The range: using 31 December as a base date, an average approach would give a
reduction in the rebate rate of 4.9% of up to 0.1% to 0.2% lower depending on the
period adopted – 0.1% being the position on a six-month average view.
Age profile
11.17 The issue: all the anecdotal data would suggest that the age profile in private sector
will be moving higher relative to public sector due to the very few new younger
members having access to a scheme in the private sector. This is due to so many
private sector schemes being closed to new members.
11.18 The evidence: the problem here is that we are unable to gain any evidence to
indicate the extent of this effect.
11.19 The range: in the absence of evidence, I feel this effect could increase the rebate by
up to 0.3% at the top end.
Remarriage (see 7.12)
11.20 Although all the consultation responses suggested to ignore this feature, it does
represent an over-statement of up to 0.05% in the resulting rebate, and so I mention
it here for completeness.
Equity returns
11.21 The respondents to the consultation were generally comfortable with the assumption
regarding future equity returns. The evidence outlined in Appendix D suggests that
equities are expected to return between 2.5% to 4.0% over cash and between 1.5%
and 4.0% over nominal bonds. I have assumed a return of 7.40% per annum which
represents a risk premium over nominal gilts of 2.75%. at the centre of this range.
11.22 This implies that any potential increase in the rebate from a lower assumption is
balanced by a potential decrease from a higher assumption, so that I have not
considered it necessary to include this assumption in the table summarising the
issues below. That said, a reasonable view would be for changes to this central rate
implying up to 0.2% change in the rebate in either direction.
30
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Summary of issues
Issue Base assumption Effect on the rebate percentage
(4.9%), if using an alternative
assumption within a best estimate
framework
CPI 2% Up to 0.3% extra
Pensions cannot fall CPI will never fall below zero Up to 0.1% extra but total impact on
rebate of this and CPI no higher than
0.3%
Date of yield determination 31/12/2010 Up to 0.2% less
Age profile Aggregates public and
private sectors
Up to 0.3% extra
Remarriage No remarriage Up to 0.05% less
Total Between 0.25% less and 0.6% extra*
* ignoring the equity return variability mentioned in paragraph 11.22 above.
11.23 It should be noted that saying that the use of 2.0% CPI is appropriate is linked to
saying that the current view of “no evidence yet to justify anything less” is a logical
approach and analytical feel can be ignored at present. This logic is not dissimilar to
saying that the current market view on yields should have no adjustment for
analytical feel. In other words, a backward looking view on yields to use an average
ought to be the same as a backward looking view on CPI median/mean and this
latter position would imply CPI higher than 2.0%.
11.24 So, for example, a selection of 2.0% CPI and a downward adjustment to the yield
seems like an inconsistency in approach.
11.25 The other point to note is that not all of the above have equal weighting so it would
be wrong to say that the best estimate is in the middle of the range.
Conclusion
11.26 If I take the base assumptions, then I feel sufficiently confident that the overall result
can be considered as a best estimate in view of all the uncertainty. This is not
because each element is right per se but because, in the round in my estimation, the
end result for the rebate feels to be in the right part of the range. Put another way,
the elements of the assumptions could be changed a bit to derive another consistent
set which delivers the same end result. Whilst it could be argued that taking one
extreme or the other of the range for each assumption might appear to be best
estimate (because each component part might be), the mathematics and logic would
not allow such a conclusion in my estimation.
31
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
11.27 In considering this part of my analysis, the Secretary of State will, of course, need to
take into account other matters. The decision on which rebate rate to use for the five
year period is not something which can or should be based solely on the actuarial
analysis and so it is hoped that the three approaches I offer can assist the Secretary
of State to decide whether to adopt a rate at one of the levels offered or indeed at a
level between the rates offered or indeed totally outside the range.
11.28 Taking the above analysis into account, my view on the best estimate rebate can be
determined from the following assumption set:
CPI (annual)
CPI (multi-year accumulation)
RPI(CPI + 0.75%)
NAE (RPI + 1.5%)
Index-linked gilt return
(RPI + real yield)
Fixed interest gilt returns
Corporate bond returns
(fixed interest + 0.75%)
Equity returns
Investment expenses
Net pre-retirement discount rate
Net post-retirement discount rate
2% per annum
2% per annum
2.75% per annum
4.25% per annum
3.35% per annum
(using 31/12/10 data)
4.65% per annum
(using 31/12/10 data)
5.40% per annum
(using 31/12/10 data)
7.40% per annum
(that is, 2.75% above gilt returns)
0.12% pa to 0.15% pa
(depending on term to SPA)
0.18% to 2.33% based on age
(see paragraph11.35 below)
2.39% (see paragraph 11.29 below)
Net post-retirement discount rate for best estimate
11.29 The net post-retirement discount rate represents the return on appropriate
investments after state pension age in excess of the annual increases on state
benefits forgone.
11.30 After allowing for the assumed asset allocation set out in paragraph 8.6 and
deducting investment expenses, the assumed return is 4.44% pa:
(30% x 3.35% + 30% x 4.65% + 40% x 5.40%) - 0.12% = 4.44%
11.31 Allowing for the CPI increase assumption of 2% pa gives a net return over prices of
2.39% calculated as:
32
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
(1 + 4.44%) / (1 + 2%) – 1 = 2.39%
Net pre-retirement discount rate for best estimate
11.32 The net pre-retirement discount rate represents the return on appropriate
investments before state pension age in excess of the revaluation of earnings
factors on benefits foregone (by reference to National Average Earnings growth).
11.33 The rates are expressed as single rates which apply for the whole period to state
pension age (SPA). However, they are derived and calculated based on an asset
allocation which varies for the individual as the individual grows older and closer to
SPA.
11.34 Based on the assumed asset allocation set out in paragraph 8.6 and deducting
investment expenses, the assumed gross returns are:
Just before SPA
[30% x 3.35% + 30% x 4.65% + 40% x 5.40%] - 0.12% = 4.44%
10 years before SPA
[50% x 7.40% + 30% x 4.00% + 20% x 5.40%] - 0.14% = 5.84%
Over 25 years to SPA
[80% x 7.40% + 12% x 4.00% + 8% x 5.40%] - 0.15% = 6.68%
11.35 Allowing for the NAE growth assumption of 4.25% gives a net of earnings
assumption of 0.18% pa for those currently just before SPA. This increases to
1.53% pa for those currently 10 years to SPA and to 2.33% pa for those currently 25
or more years from SPA. These rates are sample rates with different rates for
intermediate ages. They are calculated to apply over the whole period to SPA for
any individual of the age in question.
Evidence
11.36 See Appendix D for more detail.
33
>
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
12 Derived defined benefit contracted-out rebates
In this section I set out the defined benefit contracted-out rebates which I have derived
based on the assumption sets discussed earlier in this report.
Rebates are stated based on State pension ages as currently defined in legislation, and
also as they would be if intended changes to State pension ages are made as expected.
I also provide a reconciliation showing how the rebate values have changed since the
previous report (completed by my predecessor as Government Actuary and dated January
2006).
12.1 Based on the assumptions outlined in previous sections, the derived rebates are:
Using state pension ages based on existing legislation
4.9% on a ‘best estimate’ basis,
6.3% on a ‘typical funding’ basis, and
10.5% on a ‘gilts’ basis.
Using state pension ages based on expected revisions to existing legislation
>
4.8% on a ‘best estimate’ basis,
6.2% on a ‘typical funding’ basis, and
10.4% on a ‘gilts’ basis.
12.2 The rebate percentages are of earnings between the Lower Earnings Limit and the
Upper Accrual Point.
Reconciliation of changes to the rebate value
12.3 At the previous review, the proposed rebate of 5.8% was on a basis which included
some margins of prudence compared to a best estimate, though it was not
rationalised in the same way as the ‘typical funding’ basis proposed now. The
Secretary of State then decided to set the level at 5.3% – the same as had been
used for the previous five years. This was split 1.6% employee and 3.7% employer.
12.4 The following table shows the effect of the various factors which affect the derived
rebates since the previous review:
34
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Rebate proposed by the
Government Actuary in 2006
5.8%
Though the Secretary of State
adopted 5.3%
Change to benefits caused by
increasing year of birth
(0.3%) Accrual is falling over time (-ve)
There is a cohort effect for women
due to the SPA change to 65 (-ve)
Assumed mortality incorporates
improvements over time (+ve)
2010 rebate on 2005 basis
(before allowing for known changes)
5.5%
Change to State Pension Age
(from 65 to 68 for everyone)
(0.5%) Longer discounting period (-ve)
Lower accrual (-ve)
Alternatively, (0.6%) if we take into
account proposed further
changes.
Change to increases in payment (0.7%) Changed from RPI to CPI
Member profile 0.3% Based on recent sample data, but
including some projection
assumptions
2010 rebate on 2005 basis
(after allowing for known changes)
4.6%
Change of assumptions 1.8% financial – lower discount rates
0.4% demographic – improved longevity
(0.5%) Explicit margin in the previous
basis
Rebate in 2010 on ‘typical
funding’ basis
6.3%
Using SPAs based on existing
legislation
35
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
13 Other practical issues
In this section I address some miscellaneous practical issues.
Effect of contracting-out on staff and employers
13.1 The defined benefit contracted-out rebate is derived having regard to the average
worker contracted-out on a defined benefit basis.
13.2 For each individual worker contracted-out on a defined benefit basis, the only direct
effect of the review is if the employee portion of the rebate changes. Neither the
state second pension nor occupational pension is affected directly by the contracted-
out rebate review.
13.3 For COSRS, and their sponsoring employers, the direct effect of the review is that
the contracted-out rebate they receive may change. It does not affect the pensions
they are committed to provide. However, indirectly, the pension scheme is
effectively substituting part of the state pension. That is, part of the occupational
pension can be considered as being a replacement for the state benefits forgone, for
which the scheme (and worker) receive the rebate in compensation.
13.4 The attractiveness of contracting-out to employers depends on the balance between
the state pension forgone by contracted-out workers, and the rebate received in
compensation. Different employers will have different views on the adequacy and
attractiveness of the defined benefit contracted-out rebate.
13.5 In section 6, I outlined three alternative bases for deriving a defined benefit
contracted-out rebate, which represent a range of potential rebate values. A
rationale for each basis is given in Section 6.
13.6 Many of the consultation replies focused on the appropriate basis for calculating the
rebate. Several respondents expressed concern that a rebate calculated on a ‘best
estimate’ basis is not appropriate. Support for a ‘typical funding’ and ‘gilts’ basis
was mixed.
13.7 I do not see it as my job to take a political view on the policy objective here. So I
believe it is right to offer various bases along with analysis of what each means and
leave the political decision to Ministers.
Value of state guarantee
13.8 Several respondents to GAD’s public consultation made various comments whose
common theme was that the rebate should be based on assumptions which
recognise that the state benefits given up by contracted-out workers have a strong
(but not complete) government guarantee.
36
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
13.9 The relevant legislation
7
refers to ‘the cost of providing benefits of an actuarial value
equivalent to that of the benefits’ which are forgone by contracted-out workers. I do
not believe that this phrase requires the rebate to explicitly recognise any value of
the guarantee separate from the value of the benefits, though I accept that it would
not prohibit such an approach. I note that if Parliament had intended that the rebate
should include some additional value to recognise the nature of the guarantee
applying to state benefits, then it could have added explicit language to the
legislation to that effect.
13.10 It is also worth bearing in mind that in previous reviews, this specific point was not
allowed for and how this plays out might be arguable. So, to me, the natural
approach is to ignore this feature based on precedent and practice.
Age of membership
13.11 The rebate will be less attractive to employers with a workforce which is older on
average than the general contracted-out population, since pensions are generally
more expensive to provide for older workers.
13.12 At previous reviews, a margin was included (a loading of 7.5% to the proposed
defined benefit rebate) in order to address such issues.
13.13 The rebates derived under the three methodologies set out in this report do not
include any such margin. The approach I have adopted for this review means that a
range of possible rebates has been presented, along with the rationale underlying
them (in particular the level of prudence). It is anticipated that, in this context, the
Secretary of State can take into account issues such as cost variations between
schemes in determining the defined benefit rebate to be paid to the extent he so
wishes.
Members affected – public sector and private sector considerations
13.14 The relevant population of contracted-out workers includes both public and private
sector pension schemes.
13.15 The defined benefit contracted-out rebate values outlined in this report have been
derived using assumptions and methodologies which recognise the practice of
funded private sector pension schemes. Although the defined benefit contracted-out
rebate also applies to members of unfunded public sector schemes (who comprise a
material percentage of members contracted-out on a defined benefit basis), I have
not considered this further, since the contracting-out of public sector schemes does
not result in cash flow to or from the Government, except in relation to employee
contributions and/or non-Central Government employers.
13.16 Contracted-out public servants are affected by the level of the rebate which accrues
to employees. However, this is a decision taken by the Secretary of State on
broader policy grounds without the need for actuarial advice.
7
Pension Schemes Act 1993 section 42 paragraph (1)(a)(ii)
37
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
13.17 As I am focusing on the ‘cost’ in private sector schemes of providing the state
second pension benefits forgone, a number of consultation respondents suggested
that I also exclude public sector data on the profile of contracted-out members from
the calculation of the rebate (see the assumptions on ‘weights’ discussed from
paragraph 7.16.) There is a viable rationale for this point of view.
13.18 Conversely, there is a difference between determining the average cost using
principles which reflect private sector scheme practice, and ignoring the profile of
public sector members in the analysis used for the weights. The same defined
benefit contracted-out rebate applies to both types of schemes, and it is worthwhile
to derive rebates which reflect the total membership of contracted-out schemes, in
both sectors.
13.19 In practice, data is not available to accurately allow for each sector individually.
Thus any attempt to base the rebate only on the private sector would be speculative.
13.20 When considering the appropriate level of the defined benefit contracted-out rebate,
in the context of the alternative approaches outlined, it should be borne in mind that
reflecting the profile of all contracted-out members in the calculations may result in a
lower contracted-out rebate.
13.21 The defined benefit contracted-out rebate is only relevant to schemes that are open
to future accrual.
13.22 Notwithstanding all the above, the possible implications of the approach adopted for
a best estimate rebate rate are looked at in section 11 and considered within a
holistic framework.
38
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
14 Process for decision
In this section I offer a possible process to assist the Secretary of State reach his decision
on the appropriate level for the defined benefit contracted-out rebate.
14.1 I suggest that the Secretary of State might follow a process to reach a decision on
the appropriate level for the defined benefit rebate as follows:
>
consider the broad approach to valuing the state benefits given up by
contracted-out staff (for example, the ‘best estimate’, ‘typical funding’ and ‘gilts’
bases described in this report, or indeed any variant in between or outside the
range offered here). The Secretary of State should consider what he believes is
the appropriate policy objective in coming to this view noting (but not
necessarily taking account of, since it is not a representative cross-section of
public opinion) that many consultation responses did not support a policy
objective reflected by the use of a ‘best estimate’ approach,
consider, if appropriate, any refinements such as:
o if the rebate should include explicit allowance for the guaranteed nature of
the state pension given up by contracted-out staff (see paragraph 13.8),
o if the rebate should include a weighting to mitigate the financial effect on
schemes with older than average memberships, as recommended at
previous reviews (see paragraph 13.12),
o whether or not to allow for expected changes to state pension age,
o whether there is any further and better evidence (which would be
accepted independently as ‘further and better’) which might suggest an
assumption change but being aware of the age profile issues – see
paragraph 11.17, and
o allowance for any margins, rounding, etc.
> determining the rebate and the split between employer and employee (noting
that this aspect is outside of any considerations in this report).
>
39
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
15 Revaluation rate for Guaranteed Minimum Pensions
In this section I discuss the rate of revaluation which might be prescribed for the GMPs of
staff who leave pensionable service in the period 2012 to 2017.
15.1 I have been asked by the Department of Work and Pensions to propose a rate for
the revaluation of Guaranteed Minimum Pension (GMP) for contracted-out members
who leave pensionable service in the period 2012 to 2017.
15.2 Schemes may choose between two approaches to revaluation in this situation:
>
>
revalue GMPs in line with increases in National Average Earnings (as specified
in annual orders), or
revalue GMPs in line with a fixed rate (as specified in orders which apply for
leavers in specified five year periods).
15.3 This section deals with the rate to be determined under the second bullet point
above.
15.4 At previous reviews the then Government Actuary based his proposal for the latter
fixed rate on the assumed rate of growth in average earnings underlying his
recommended rebate value for defined benefit members, plus 0.5%. The addition of
0.5% to the expected future growth in average salaries is to compensate for the
conversion of a variable rate of revaluation (linked to actual average earnings
growth) to a fixed rate for which there has to be a ‘price’.
15.5 I propose to continue to adopt this rationale. However, at this review I have
provided a range of three possible rebate values, which are based (among other
things) on differing assumptions in respect of average earnings growth. In my
opinion, a best estimate assumption is most appropriate for the current purpose.
15.6 Adding 0.5% in line with previous practice, I therefore propose that the rate for the
revaluation of GMP for members who leave pensionable service in the period 2012
to 2017 should be set at 4.75%. This compares to a current rate of 4.0%.
15.7 I recommend that this revaluation rate would be appropriate in the situation where
the National Insurance rebate was set at a rate of 6% or less.
15.8 If a higher rebate were set, then this would imply that state benefits were considered
to be more valuable than implied by my ‘best estimate’ bases. This might reflect a
view that National Average Earnings were expected to rise more quickly than
allowed for in my ‘best estimate’ basis. If so, that would justify a higher revaluation
rate. Therefore, if the National Insurance rebate were set at a rate over 6% then I
suggest that the Secretary of State should consider setting the GMP revaluation rate
at 5.5%.
40
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Appendix A Legal background
State second pension and contracting-out
A.1 The state second pension is set out in the Social Security Contributions and Benefits
Act 1992 (SSCBA 1992). The Category A pension (paid to the contributor) is
described from s44. The Category B pension (inherited by the spouse of the
contributor) is described from s48A. The amounts of the state second pension (and
the effect of contracting-out) are set out in Schedule 4A in respect of accrual after
2002/03 and before the ‘flat rate introduction year’. Accrual from the ‘flat rate
introduction year’ is set out in Schedule 4B. The ‘flat rate introduction year’ has not
been prescribed yet, but it is intended that it will be 2012/13.
A.2 The amount of the state second pension forgone by those contracted-out on a defined
benefit basis is defined identically in both Schedule 4A and 4B, so no difficulty arises
in this regard from the uncertainty around the definition of the ‘flat rate introduction
year’.
A.3 For those contracted-out on a defined contribution basis, only Schedule 4A provides
definitions of contracted-out benefits, since the ‘flat rate introduction year’ is expected
to be synchronised with the abolition of contracting-out on a defined contribution
basis.
National Insurance rebates
A.4 The Pension Schemes Act 1993 (PSA 1993), sections 40 to 49 inclusive, makes
provision for members of pension schemes who are contracted-out of the state
second pension (and the sponsoring employers of those schemes) to pay reduced
rates of National Insurance contributions and/or to have their pension schemes
receive corresponding payments from HM Revenue & Customs (HMRC).
A.5 PSA 1993 s42 requires a review to be carried out by the Government Actuary, at
intervals of not more than five years, of the contracted-out rebate for those
contracted-out on a defined benefit basis (members of COSRS). The review should
report on any changes in the factors which, in the Government Actuary’s opinion,
affect the cost of providing benefits of an actuarial value equivalent to that of the state
pension benefits which are given up by or in respect of members of these schemes.
The legislation does not require the Government Actuary to recommend percentage
rebates, though in practice the Department for Work and Pensions commissions GAD
to provide advice on the level of rebate values (and there is no legislative bar
preventing this).
A.6 Separate requirements exist for members of Contracted-Out Money Purchase
Schemes (COMPS) under PSA 1993 s42B. COMPS are occupational defined
contribution schemes. Contracted-out rebates are age-related for members of
COMPS. The Government Actuary must report on the percentage age-related
rebates which are required in his opinion to reflect the cost of providing benefits of an
actuarial value equivalent to that of the state pension benefits which are given up by
or in respect of members of COMPS.
41
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
A.7 PSA 1993 s45A specifies age related rebates for Appropriate Personal Pensions
(APPs). These rebates are paid by HMRC direct to the schemes. The Government
Actuary must report on the age-related rebates which are required in his opinion to
reflect the cost of providing benefits of an actuarial value equivalent to that of the
state pension benefits which are given up by or in respect of members of APPs.
A.8 The same sections of PSA 1993 require any order by the Secretary of State for Work
and Pensions to vary the rate of reduction in National Insurance contributions or
rebates, to which the Government Actuary’s reports would relate, to be made at least
one complete tax year before they come into force and to be laid within 5 years of the
previous reports and draft order. Hence for new rebates to apply from 6 April 2012,
the relevant orders and the Government Actuary’s reports must be laid before
Parliament by 28 February 2011, since the previous order was laid on 1 March 2006.
Changes since the previous review
A.9 For people in contracted-out employment, section 10(3)(b) of the Pensions Act 2007
(PA 2007) provides that the state second pension moves to accrual on two bands
(previously there were three bands) from 2010/11.
A.10 Section 15(1) of PA 2007 provides that contracting-out on a defined contribution basis
(via COMPS and APPs) will be abolished from a date appointed by the Secretary of
State. The abolition date has not yet been set, but is expected to be April 2012, at the
same time as the Band 1 accrual of the state second pension moves to a flat rate.
A.11 Section 13 of PA 2007 provides that state pension age will rise from 65 to 68 over the
period 2024 to 2046.
A.12 At the June 2010 budget, the Chancellor of the Exchequer announced that state
second pension will be increased in payment in line with the Consumer Prices Index
rather than the Retail Prices Index.
A.13 As part of the 2010 spending review, the Government announced plans to accelerate
the rise in state pension ages. Subject to Parliamentary approval the state pension
age should rise to age 65 for all by 2018 and age 66 by 2020.
Reduced state pension
A.14 The effect of contracting-out on a member’s state second pension is set out in the
Social Security Contributions and Benefits Act 1992 (SSCBA 1992) Schedules 4A
and 4B.
A.15 For those contracted-out on a defined benefit basis, members give up a percentage of
their earnings between the Qualifying Earnings Factor (that is, the Lower Earnings
Limit expressed as an annual amount) and the Upper Accrual Point (expressed as an
annual amount).
A.16 The percentage is 20 divided by the number of ‘relevant years’ in the member’s
‘working life’. ‘Working life’ (defined in SSCBA 1992 schedule 3 paragraph 5(8)) is the
period between the tax year the member attains the age of 16, and the tax year
immediately before attaining state pension age. ‘Relevant years’ (defined in SSCBA
1992 s44(7)) are restricted to 1978/79 and later.
A.17 The benefits forgone would otherwise have been payable at state pension age.
42
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
A.18 From the ‘flat rate introduction year’ the state second pension itself is amended. For a
few members (older and higher earning) the state second pension is reduced to zero.
That is, the specified deduction for contracted-out workers exceeds their state second
pension.
Inherited benefits (simplified summary)
A.19 Where both the contributor and the spouse (including civil partners) are over state
pension age at the date of the contributor’s death then the spouse generally becomes
entitled to a pension equal to half of the contributor’s accrued state second pension
(though in some cases inheritance is capped at a maximum value).
A.20 If the spouse is under state pension age then the inherited pension is not paid until
state pension age unless there are dependent children. If the spouse is between 45
and 55 and there are no dependent children then the inherited pension is reduced, as
well as being delayed until state pension age. If the spouse is under 45 and there are
no dependent children then the inherited pension is eliminated.
Reduced National Insurance Contributions
A.21 The Pension Schemes Act 1993 s41(1), (1A) and (1B) provide that the Class 1
National Insurance contributions payable in respect of the member contracted-out on
a defined benefit basis should be reduced by a percentage of their earnings between
the Lower Earnings Limit and the Upper Accrual Point.
Defined contribution schemes
A.22 For the time being, COMPS have the same state pension reduction as COSRS.
Members of APPs generally forgo their entire entitlement to state second pension.
A.23 The National Insurance rebates for those contracted-out on a defined contribution
basis are age related, and may differ between COMPS and APPs.
43
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Appendix B Assumptions adopted at the previous review
The following assumptions were adopted at the previous review, in respect of the defined
benefit contracted-out rebate.
Financial
Pre-retirement discount rate (net of earnings
growth)
1% pa just before SPA
2% pa 15 yrs to SPA
2.5% pa >30 yrs to SPA
Post-retirement discount rate (net of pension
increases)
2.00% pa
Gross pre-retirement discount rate 5.00% pa
Demographic
Mortality
85% of the projected mortality
used for the UK 2004-based
national population projections.
The same proportion of
population mortality was
assumed for the widows and
widowers of members.
Proportions ‘married’
In line with the 2003-based national
projections.
Rates of ‘remarriage’
In line with the 2003-based national
projections.
‘Marital’ age differences
In line with the 2001 national census
data.
Weights
Derived from data analysed by
the Department of Work and
Pensions.
Administration expenses 0.2% of band earnings
Contingency margin 7.5% of calculated rebate
44
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Appendix C Assumptions adopted at this review
The following tables summarise the assumptions used to provide illustrative defined benefit
contracted-out rebates for each of the three approaches discussed. The ‘building blocks’ of
our assumptions have been rounded to 0.05% pa.
‘Best estimate’ basis
Financial
CPI (multi-year) 2.00% pa
RPI (CPI + 0.75%) 2.75% pa
NAE (RPI + 1.50%) 4.25% pa
Index linked gilt return (RPI + real yield) 3.35% pa
Fixed interest gilt returns 4.65% pa
Corporate bond returns (Fixed interest gilts + 0.75%) 5.40% pa
Equity returns 7.40% pa
Investment expenses (depending on term to SPA) 0.12% pa to 0.15% pa
Pre-retirement discount rate (net of earnings growth)
0.18% pa just before SPA
1.53% pa 10 years to SPA
2.33% pa >25 yrs to SPA
Post-retirement discount rate (net of pension
increases)
2.39% pa
Gross pre-retirement discount rate 4.44% pa
Demographic
Mortality
Based on S1 tables, allowing for
mortality improvements in line
with the ONS principal Uk
projections of population (2008
series).
Proportions ‘married’
In line with the 2006-based national
projections.
Rates of ‘remarriage’ None
‘Marital’ age differences
In line with the 2001 national census
data.
Other
Weights
Derived by GAD from National
Insurance sample data
Administration expenses 0.2% of band earnings
Contingency margin None
45
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
‘Typical funding’ basis
Financial
CPI (multi-year) 3.00% pa
RPI (CPI + 0.75%) 3.75% pa
NAE (RPI + 1.50%) 5.25% pa
Index linked gilt return (RPI + real yield) n/a
Fixed interest gilt returns 4.65% pa
Corporate bond returns (Fixed interest gilts +
0.75%)
n/a
Equity returns n/a
Investment expenses (depending on term to SPA) Implicit in discount rates
Pre-retirement discount rate (net of earnings growth) 1.19% pa
Post-retirement discount rate (net of pension
increases)
1.94% pa
Gross pre-retirement discount rate 5.00% pa
Demographic
Mortality
Based on S1 tables, allowing
for mortality improvements in
line with the ONS principal UK
projections of population (2008
series). Rates are offset by two
years of age.
Proportions ‘married’
In line with the 2006-based national
projections.
Rates of ‘remarriage’ None
‘Marital’ age differences
In line with the 2001 national census
data.
Other
Weights
Derived by GAD from National
Insurance sample data
Administration expenses 0.2% of band earnings
Contingency margin None
46
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
‘Gilts’ basis
Financial
CPI (multi-year) 3.25% pa
RPI (CPI + 0.75%) 4.00% pa
NAE (RPI + 1.50%) 5.50% pa
Index linked gilt return (RPI + real yield) n/a
Fixed interest gilt returns 4.65% pa
Corporate bond returns (Fixed interest gilts +
0.75%)
n/a
Equity returns n/a
Investment expenses (depending on term to SPA)
0.10% pa
Pre-retirement discount rate (net of earnings growth) - 0.90% pa
Post-retirement discount rate (net of pension
increases)
1.26% pa
Gross pre-retirement discount rate 4.55% pa
Demographic
Mortality
Based on S1 tables, allowing
for mortality improvements in
line with the ONS principal UK
projections of population (2008
series). Rates are offset by two
years of age.
Proportions ‘married’
In line with the 2006-based national
projections.
Rates of ‘remarriage’ None
‘Marital’ age differences
In line with the 2001 national census
data.
Other
Weights
Derived by GAD from National
Insurance sample data
Administration expenses 0.2% of band earnings
Contingency margin None
47
>
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Appendix D Evidence in support of the financial assumptions
In this appendix I present and discuss some of the published evidence which is relevant to
the choice of financial assumptions. This includes:
evidence from Pension Regulator reports in respect of discount rates
adopted by funded pension schemes,
evidence from Pension Regulator reports and NAPF reports on the asset
allocation strategy adopted by funded pension schemes,
research on expected prices and salary inflation,
research on expected asset returns,
comments on expected investment management expenses.
‘T
ypical funding’ basis considerations – evidence from Pension Regulator reports
D.1 The Pensions Regulator’s report ‘Scheme funding: an analysis of recovery plans’,
dated November 2009, provides an overview of the first triennial cycle of the new
scheme funding regime for defined benefit and hybrid pension schemes. The triennial
cycle is analysed in three tranches over the period from September 2005 to
September 2008.
D.2 The report states:
‘The discount rate can be broadly described by the following equation:
discount rate = risk free rate + risk premium
A proxy such as a government bond yield is typically used for the risk free rate,
and a spread (i.e. a risk premium) over the risk free rate is assumed, typically
based on:
• The time horizon of the liabilities
• The potential for additional investment return; and
• A prudence adjustment, based on the employer’s covenant’
D.3 The table below sets out the spread over the gilt yield of the weighted average
discount rate data for those schemes that adopt different discount rates pre and post
retirement.
48
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Tranche 1
8
Tranche 2 Tranche 3
Excess of pre retirement discount rate
9
over gilt yield
2.0% pa 1.6% pa 1.9% pa
Excess of post retirement discount rate
over gilt yield
0.3% pa 0.2% pa 0.6% pa
Source: tPR and GAD
D.4 The average pre retirement spread over the period is 1.83% pa and the average post
retirement spread over the period is 0.37%.
D.5 The report notes that:
‘there has been an increase in the discount rate spread over UK gilt yields. This
trend may be because schemes are adopting an increased risk premium when
setting discount rates. This increase…in tranche 3 implies all else being equal, a
slightly greater reliance on investment outperformance to meet scheme liabilities.’
Furthermore, the report refers to this ‘investment outperformance’ as
corresponding to the increase in corporate bond spread over gilts and states that:
‘anecdotal evidence from case work by the regulator suggests that some schemes
are using corporate bond yields at least in part as a basis for setting the discount
rate’.
D.6 We consider it appropriate to base our assumptions for the pre and post retirement
margins above gilt yields used for setting the typical funding discount rates on the
averages given in D.3.
Pension scheme asset allocation – NAPF and Pension Regulator evidence
D.7 The NAPF 2009 Annual Survey provides statistics on defined benefit pension
schemes with assets to the value of approximately £400 billion (and tends to cover
larger schemes). Notably, page 29 sets out the asset allocation of the defined benefit
schemes covered by the survey. Approximately 34% of assets are shown as
government and corporate bonds, of which around 15% are shown as the latter. This
suggests that around 44% of the ‘bond’ assets held by the pension schemes covered
by the survey were corporate bonds.
D.8 In respect of the split of assets between the UK and Overseas, the NAPF 2009 survey
shows that the split of equities held was approximately 50:50. The vast majority of
government bonds held were UK gilts.
8
The tranches refer to each of the three years between September 2005 and September 2008.
9
Where ‘pre and post retirement discount rates’ are before deduction of an inflation assumption, and
gilt yield is measured as the yield on the over 15 year UK gilt index
49
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
D.9 The 2009 Purple Book published by the Pensions Regulator covers almost all pension
schemes in the PPF eligible UK defined benefit scheme universe. Section 7.6 (Chart
7.4) of the Purple Book shows that, broadly speaking:
schemes with around 90% non pensioner and 10% pensioner liabilities have an
average asset allocation of approximately 65% equities and property, 30% gilts
and fixed interest and 5% cash and other,
schemes with around 50% non pensioner and 50% pensioner liabilities have an
average asset allocation of approximately 50% equities and property, 40% gilts
and fixed interest and 10% cash and other, and
schemes with around 10% non pensioner and 90% pensioner liabilities have an
average asset allocation of approximately 20% equities and property, 70% gilts
and fixed interest and 10% cash and other
D.10 Section 7.3 of the Purple Book shows that the weighted average allocation of ‘bond’
assets to government bonds was 29%, to index-linked bonds was 32.6% and to
corporate bonds was 38.3%.
Consumer Prices Index (CPI) and Retail Prices Index (RPI)
D.11 The CPI and RPI measures of inflation differ in both their construction methods and in
the items which are included. The formula for CPI uses a geometric mean to combine
prices within each category. This gives a lower mean than the arithmetic method that
is used for RPI. The Office for National Statistics (2003)
10
calculated that this formula
effect meant that RPIX (which uses the same construction method as RPI) was about
0.5% greater than CPI for historic data, with an additional difference of 0.2% pa due
to differences in items included. Each month, the Office of National Statistics
publishes the breakdown of the difference between RPI and CPI which has shown the
same formula effect of about 0.5%.
D.12 Since 1992, when inflation targeting began, RPI has been 0.7% pa greater than CPI
11
.
HM Treasury has often used a difference of 0.75% pa for their budget projections, for
example HM Treasury (2007)
12
, and this is also consistent with the views of King
(2007)
13
who said the Bank of England expected an average long run gap of about
0.7% / 0.8% pa.
D.13 The Office for Budget Responsibility’s expectations of the gap between CPI and RPI
in the short-term are much higher at 1.0% -1.5%. This is likely to be due to
expectations that rises in interest rates will push up RPI relative to CPI (as unlike CPI,
RPI includes mortgage interest payments).
10
Office for National Statistics (2003) ‘The New Inflation Target: the Statistical Perspective’
11
The difference in geometric means between October 1992 and March 2010 using Office for National
Statistics data
12
HM Treasury (2007) ‘Budget 2007’
13
King, Mervyn (2007) ‘Inflation report press conference – 16 May 2007’
50
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
D.14 Over the period 1900 to 2000 Dimson, Marsh and Staunton (2002)
14
found that the
geometric average inflation has been 4.1% pa in the UK and 3.2% pa in the US.
However, the average inflation rate is very dependent on the period chosen. Since
October 1992, when the UK adopted an inflation targeting regime, inflation has been
lower. Between October 1992 and March 2010 the geometric average of RPI inflation
has been 2.7% pa
15
.
D.15 An alternative method of estimating future inflation is to look at the inflation required
so that the return on index linked bonds is the same as that on nominal bonds.
UK instantaneous implied inflation forward curve
Source: Bank of England
D.16 However such breakeven rates may also contain risk premia and these must be
removed to calculate the expected inflation. Generally, it is considered that the risk
premium for nominal bonds is positive which is consistent with investors demanding
an additional expected return to compensate them for the risk that their real return is
eroded by periods of high inflation. The inflation risk premium is difficult to estimate
and may vary by term and through time. A good summary of studies investigating the
size of the inflation risk premium can be found in Hördahl (2008)
16
, which has been
quoted below.
14
Dimson, Marsh and Stauton (2002) ‘Triumph of optimists, 101 years of global investment returns’
15
Calculated using data from the Office for National Statistics
16
Hördahl (2008)
16
‘The inflation risk premium in the term structure of interest rates’, BIS Quarterly
Review September 2008
51
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
‘The available empirical evidence on the properties of inflation risk premia is
somewhat mixed. Studies that cover very long sample periods and that do not
include information from index-linked bonds to help pin down the dynamics of real
yields often report sizeable inflation risk premia. For example, using a structural
economic model, Buraschi and Jiltsov (2005) find that the 10-year US inflation risk
premium averaged 70 basis points from 1960.
17
They also find that the inflation
premium was highly time-varying, and that by the end of their sample it had fallen
to relatively low levels. Ang et al (2008) estimate a term structure model in which
inflation exhibits regime switching using US inflation and nominal yield data, and
report a large and time-varying inflation risk premium (on average, around 115
basis points for the five-year maturity over their 1952–2004 sample).
In papers that focus on more recent periods and in those that utilise information
embedded in index-linked bonds, inflation risk premium estimates tend to be
relatively small, although still mostly positive. Durham (2006) estimates a no-
arbitrage model using US Treasury inflation-indexed bond data and finds that the
10-year inflation premium hovered around a slightly positive mean from 2003
onwards.
18
D’Amico et al (2008) apply a similar model to data from 1990 onwards,
and report a moderate-sized positive 10-year inflation premium (around 50 basis
points on average) that is relatively stable. However, they also find that their
results are sensitive to the choice of date from which index-linked bond data are
included.
The available empirical evidence relating to euro area data is more limited. In fact,
apart from the papers on which this article is based, there appears to be only one
study focusing on the euro area.
19
García and Werner (2008) apply a term
structure model similar to that used by D’Amico et al (2008) on euro real and
nominal yields, supplemented with survey data on inflation expectations. Their
estimates suggest that the inflation premium at the five-year horizon has averaged
around 25 basis points since the introduction of the euro, and that it has fluctuated
only mildly over time. Hence, their results seem to be in line with those of Durham
(2006) and D’Amico et al (2008), which point to a relatively modest, but positive,
long-term inflation risk premium in recent years.’
17
All quantitative risk premium estimates mentioned are in terms of (annualised) yield, rather than for
example holding period returns.
18
Prior to 2003, Durham (2006) obtains a 10-year inflation premium that was mostly negative. This is
probably due to sizeable liquidity premia in this part of the sample period, which would have tended to
raise the index-linked bond yield and therefore produce negative inflation premia to fit the resulting low
level of break-even inflation rates.
19
Prior to 2003, Durham (2006) obtains a 10-year inflation premium that was mostly negative. This is
probably due to sizeable liquidity premia in this part of the sample period, which would have tended to
raise the index-linked bond yield and therefore produce negative inflation premia to fit the resulting low
level of break-even inflation rates.
52
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
D.17 Hördahl (2008) estimates the inflation risk premium using a dynamic term structure
model based on an explicit structural macroeconomic model. Using this model he
estimated that in 2008 the inflation risk premium on 10-year bonds in the Euro area
was about 0.5% pa but that it was lower in the US (about 0.1% pa), as illustrated by
the graphs
20
below:
Source: Hördahl (2008)
D.18 PricewaterhouseCoopers (2007)
21
reviewed investment return projection rates for the
FSA, and suggested an appropriate assumption of 2.75% pa for RPI assuming that
the inflation targeting regime by the Bank of England continued. In coming to this
conclusion they considered both the breakeven inflation (which was then about 3.15%
pa) as well as the Bank of England’s target.
D.19 The forward breakeven RPI inflation rate
22
between 4 and 44 years time is 4.00% pa.
An assumed future gap between RPI and CPI of 0.75% implies using a forward CPI
rate of 3.25% pa. The Bank of England produces regular forecasts for future CPI over
a three year period, based around the mode of the distribution. This is part of the
process by which the 2% rate is targeted. They also estimate the mean and median
of this short-term distribution. However, the Bank does not publish any longer-term
distributions or any views on how the median and mean of the longer-term distribution
would relate to the mode.
20
Hördahl (2008) ‘The inflation risk premium in the term structure of interest rates’, BIS Quarterly
Review September 2008
21
PricewaterhouseCoopers (2007) ‘Review of FSA Projection Rates’
22
Calculated using the Bank of England nominal gilt forward curves from 31 December 2010 and
extrapolating beyond the 25 years that these are available for, by assuming that the forward rate is
constant.
53
>
>
>
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
D.20 At first sight it may appear that the perspective of investors which leads to such a
breakeven inflation rate is incompatible with the perspective of economic
forecasters who believe that the 2.0% pa Bank of England target for CPI provides an
appropriate best estimate for future inflation. However, this is not necessarily the
case as the following argument demonstrates.
Beyond the immediate term, the best estimate of CPI inflation in any one year is
2.0%, in line with the Bank of England target and independent forecasts.
In practice there will inevitably be variation around the target level (that is, there
is a distribution of possible outturns each year).
When carrying out inflation modelling, the Bank of England’s central forecast is
the mode of their distribution. This implies that the most likely outturn for CPI in
any year is 2.0%.
However, the magnitude of possible deviations from the target level may be
greater on the upside than the downside, as has been demonstrated by past
‘inflation shocks’. Hence the distribution of annual inflation may be positively
skewed and so may have a mean greater than 2.0%. Whilst there is unlikely to
be consensus on the shape of this forward-looking distribution, a reasonable
assessment of the distribution might have a mean in the range of 2.0% to 2.5%.
Long-term investors are naturally concerned about the mean outcome for
inflation over the term of their investment and not just year by year. Hence they
are interested in the multi-year distribution for inflation and not just the single-
year distribution. This is complicated by the fact that there is likely to be some
positive serial correlation in annual inflation figures as a future inflation shock
may take more than one year to eliminate. Combined with any skew in the
single-year distribution, this means that the median annualised inflation figure
from the multi-year distribution may exceed the median inflation figure from the
single-year distribution. For example I have analysed the impact of
compounding one reasonable single-year distribution with a median of 2.0%
and mean of 2.5% and found the resulting multi-year distribution over 10 or 20
years to have both a median and a mean of around 2.5%.
Hence the ‘best estimate’ CPI assumption relevant to investors and consistent
with the Bank of England target may be in the range 2.0% pa to 2.5% pa.
This implies a 0.75% inflation risk premium in the gilt forward curves which is
within the range of empirical estimates quoted above.
D.21 Since July 1997 (when the Bank of England gained independence) the mean
percentage change in CPI over the previous 12 months is 1.9% pa. The
corresponding median is 1.7%.
D.22 Since December 2003 (when CPI targeting commenced) the mean percentage
change in CPI over the previous 12 months is 2.5% pa. The corresponding median is
2.4%
54
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
National Average Earnings (NAE)
D.23 There is evidence to suggest that earnings increase faster than prices over the long
term. However, in the short term there can be large fluctuations which may be driven
by business cycle fluctuations.
D.24 Since 1970 wages have increased by 1.7% pa more than prices
23
. However, the rate
of real earnings growth appears to have fallen in recent years and has averaged 1.1%
pa since the introduction of inflation targeting in 1992.
D.25 PricewaterhouseCoopers (2007)
24
carried out statistical analysis and recommended a
1.5% pa real earnings growth was used for the projection of investment illustrations.
However, they also noted a number of uncertainties surrounding this and estimated
the plausible range of real earnings growth to be around 1.0% to 2.0% pa.
Gilt returns
D.26 The expected return on fixed interest gilts bought in 4 years’ time will be equal to the
yield on gilts at that point in time. The market consistent expectation of this is the
forward rate on gilts in 4 years’ time for the term of the gilt bought. Considering 40
year gilts, the forward nominal rate between 4 and 44 years’ time on fixed interest
gilts is about 4.66% pa
25
.
D.27 Similarly, the expected return on index-linked gilts bought in 4 years time will be equal
to the yield on gilts at that point in time. Again, the market consistent expectation of
this is the forward rate on gilts in 4 years time for the term of the gilt bought.
Considering 40 year gilts, the forward real rate between 4 and 44 years time on index
linked gilts is about 0.62% pa
26
.
Corporate bond returns
D.28 Corporate bonds are exposed to the risk of default and so the expected return is
lower than their ‘promised' gross redemption yield. Part of the spread between
government and corporate bonds is due to the expected loss due to default, part
reflects a risk premium for the uncertainty in return and the residual is a non-credit
related premium (for example, to compensate for lower liquidity). Dimson, Marsh and
Staunton (2002)
27
found that the geometric mean return of US high grade corporate
bonds between 1900 and 2000 was 2.11% pa which was 0.48% pa higher than
government bonds. They then went on to comment:
23
This is the difference in the geometric averages of RPI and the UK earnings index between January
1970 and February 2010.
24
PricewaterhouseCoopers (2007): ‘Review of FSA Projection Rate’
25
This has been calculated using the Bank of England nominal gilt forward rate curves from
31 December 2010 and extrapolating beyond the 25 years that these are available for by assuming
that the forward rate is constant.
26
This has been calculated by using Bank of England nominal gilt forward rate curves from 31
December 2010 and extrapolating beyond the 25 years that these are available for by assuming that
the forward rate is constant.
27
Dimson, Marsh and Staunton (2002) ‘Triumph of the optimists, 101 years of global investment
returns’
55
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
‘High-grade corporates typically trade on redemption yields about one percentage
point higher than on government bonds. This suggests that about half the
‘promised’ yield differential fails to materialize because of defaults, downgrades,
and early calls, while around half represents the achieved risk premium.’
D.29 PricewaterhouseCoopers (2007)
28
used two methods to estimate the expected return
on corporate bonds. Their first based on the credit default swap market estimated an
expected return of 0.75% pa above government bonds for bonds rated higher than
BBB. Their second method was to use the principles underlying CAPM and multiplied
an empirical debt beta by their equity risk premium of 3% - 4% pa to produce
expected additional returns in the range of 0.3% - 0.8% pa, on top of which there was
the possibility of a liquidity premium.
D.30 Historical expected defaults and recovery rates can be found in documents issued by
ratings agencies such as Standard and Poor’s (2010)
29
and Moody’s (2010)
30
. These
can then be used to estimate future expected losses and hence expected returns on
corporate bonds.
D.31 The Bank of England use a structural credit risk model calibrated to historical default
frequencies to separate the credit spread into its constituent parts. This model was
the result of studies by Churm and Panigirtzoglou (2005)
31
and Bank of England
(2007)
32
. This model uses option pricing methodology in a similar way to the model
by Merton (1974)
33
to value the payoff to equity and bond holders. The results from
this model are illustrated in the following Bank of England (2009) graph.
28
PricewaterhouseCoopers (2007) ‘Review of FSA Projection Rates’
29
Standard and poor’s (2010) ‘Default, Transition and Recover: 2009 Global Corporate Default Study
and Ratings Transitions’
30
Moody’s (2010) ‘Corporate Default and Recovery Rate 1920 – 2009’
31
Churm and Panigirtzoglou (2005) ‘Decomposing credit spreads’ Bank of England Working Paper no.
253
32
Bank of England (2009) ‘Financial Stability Report June 2009’
33
Merton (1974) ‘On the pricing of corporate debt: the risk structure of interest rates’
56
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
D.32 In setting the expected corporate bond return assumptions it is important to also
consider current corporate bond spreads which were 1.3% for AA bonds as at
30 November 2010
34
.
Equity Returns
D.33 The table below shows the return on equities, bonds and bills, together with the rate
of inflation and GDP (where readily available) over 1900 to 2000 and 1900 to 2009:
UK US World
Real Nominal Real Nominal Real Nominal
Returns
% pa
1900 - 2000
Equities 5.8 10.1 6.7 10.1 5.8 9.2
Bonds 1.3 5.4 1.6 4.8 1.2 4.4
Bills 1.0 5.1 0.9 4.1 0.9 4.1
(2)
Inflation 4.1 3.2 3.2
(2)
UK US World
Real Nominal Real Nominal Real Nominal
Returns
% pa
1900 - 2009
Equities 5.3 9.4 6.2 9.3 5.4 8.6
Bonds 1.3 5.3 1.9 5.0 1.7 4.7
Bills 1.0 5.0 0.9 3.9 0.9 3.9
Inflation 3.9
(1)
2.9
(1)
3.0
(1)
GDP
2.1
p/capita
D.34 The table below shows the excess return on equities over each of bonds, bills, GDP
and inflation (where readily available), from 1900 to 2000 and 1900 to 2009:
UK US World
Excess of
Over bonds 4.4 5.0 4.6
Equity Return
(1)
1900-2000
Over bills 4.8 5.8 4.9
Over inflation 5.8 6.7 5.8
34
The option adjusted spread on BofA Merrill Lynch AA Sterling Corporate Index (UR28) was 128bps
at at 30/11/2010
57
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
UK US World
Excess of
Over bonds 3.9 4.2 3.7
Equity Return
(1)
1900-2009
Over bills 4.2 5.2 4.4
Over inflation 5.3 6.2 5.4
Sources:
Dimson, Marsh and Staunton: ‘Triumph of the Optimists: 101 Years of Global Investment Returns’,
Princeton University Press
Dimson, Marsh, Staunton and Wilmot: ‘Credit Suisse Global Investment Returns Yearbook 2010’
(1) calculated by GAD using information from the above sources
(2) US data
D.35 The Equity Risk Premium (ERP) is the excess of the expected return over the ‘risk
free’ rate.
D.36 When assessing the ‘risk-free’ rate, the option of using the return on bonds or bills
needs to be considered. Dimson, Marsh and Staunton (2002)
35
state ‘of these two
only treasury bills can be considered risk free.’
D.37 The historic information above shows that UK and US equities have returned 4.2%
and 5.2% pa above their respective country treasury bills, whereas ‘world’ equities
have returned 4.4% pa over ‘US’ treasury bills.
D.38 The ‘world’ estimate is suggested by PricewaterhouseCoopers (2003)
36
as the most
appropriate measure when providing an estimate of prospective ERP for UK retail
investors.
D.39 The historical returns may not give a good estimate of the current ERP required by
investors and this causes much debate about what the prospective ERP really is.
Many (such as Globob and Bishop (1997)
37
, Siegel (1999)
38
, Cornell (1999)
39
,
Dimson, Marsh and Staunton (2002), Brigham, Eugene and Ehrhardt (2002)
40
)
suggest that future returns are likely to be lower than in the past.
D.40 The two main reasons for this are summarised by PricewaterhouseCoopers (2003):
‘(a) Many markets will simply have performed better than investors expected in
the past. Observed returns will be larger than those investors actually required to
justify them investing in equities. This is the case for the US in particular, which in
the 20th century experienced a sustained period of political stability and economic
growth.
35
Dimson, Marsh and Staunton (2002) ‘Triumph of the optimists, 101 years of global investment
returns’
36
PricewaterhouseCoopers (2003) ‘Rates of Return for FSA prescribed projections’
37
Globob and Bishop (1997) ‘What long-run returns can investors expect from the stock market?’
38
Seigel (1999) ‘The Shrinking Equity Risk Premium’
39
Cornell (1999) ‘The Equity Risk Premium’
40
Brigham, Eugene and Ehrhadt (2002) ‘Financial Management’
58
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
(b) The ERP will have fallen over the historical period as equity markets became
more diversified and efficient while investors’ confidence in the future grew. The
result of this would be a significant re-rating of equities upward, which is unlikely
to be repeated in future.’
D.41 I have summarised estimates of the prospective ERP in the table below from a
number of studies.
Dimson,
Marsh and
Staunton
(2002)
Carhart and
Winkelmann
(2003)
41
PWC
(2007)
42
Graham
and
Harvey
(2009)
43
Fama and
French
(2002)
44
Siegel
(1999)
Prospective
ERP
Estimate
2.5% to 4%
over cash
3%* over US
treasury
bonds
3% to 4%
over bonds
3.46%*
over US
treasury
bonds
2.55% to
4.32% over
6 month
commercial
paper
Less
than
1.5% to
2.5%
over
bonds
* US ERP estimate only
Investment expenses
D.42 If no allowance is made for active management outperformance in assumed asset
class returns then it is consistent to consider passive investment management fees
when setting investment expenses assumptions.
D.43 The fees in the wholesale passive investment management space are very
competitive and therefore there is usually little observed difference between the
various managers. The typical fees are summarised in the table below:
Asset Class Annual Passive Management Fee
UK Equities 0.05% to 0.1%* pa
Overseas Equities 0.13% to 0.22%* pa
UK Gilts (FI or IL) 0.03% to 0.1%* pa
UK Corporate Bonds 0.08% to 0.15%* pa
* depending on size of funds under management
D.44 It should be noted that the Pensions Regulator scheme funding analysis dated
November 2009 reports on additions over the risk free rate that are net of expenses.
41
Carhart and Winkelmann (2003) ‘The Equity Risk Premium, Modern Investment Management’
42
PricewaterhouseCoopers (2007) ‘Review of FSA Projection Rates’
43
Graham and Harvey (2009) ‘The Equity Risk Premium amid a Global Financial Crisis’
44
Fama and French (2002) ‘‘The Equity Premium’
59
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Appendix E Evidence in support of the mortality assumptions
E.1 The Government Actuary’s Department has derived ‘best estimate’ mortality
assumptions based on the experience of the contracted-out workforce. These can be
expressed by reference to standard tables as follows:
Base table
45
Male mortality
S1PML
Female mortality
S1PFL
Mortality of spouses of men
S1DFL
Mortality of spouses of women
S1PML
Note that there is no table S1DML produced for widowers.
E.2 In order to determine the ‘best estimate’ assumptions, I compared mortality rates
derived from experience data representing a 1% sample of the National Insurance
Fund over the period 2000/01 to 2007/08 to standard mortality tables. The graphs
below show that the data were a good fit to the tables listed above. The experience
of women over 80 is erratic because it is based on limited data.
45
The base tables are available on the Institute and Faculty of Actuaries’ website.
60
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
E.3 Since the previous review, these SAPS tables have become established in general
use. These tables are based on the experience of defined benefit pension schemes,
so it is perhaps unsurprising that they show a good fit with the National Insurance
Fund contracted-out experience.
E.4 The state second pension forgone is accrued up to the Upper Accrual Point only. Any
amounts effect (the higher longevity exhibited by those receiving higher incomes) for
contracted-out workers is limited by this cap on band earnings.
Relevant information from Pension Regulator’s guidance and reports
E.5 In relation to scheme funding, the Pensions Regulator’s guidance highlights the
importance of a prudent approach to setting mortality assumptions for both the base
table and future improvements. The Pensions Regulator’s report, ‘Scheme Funding:
an analysis of recovery plans’ dated November 2009, indicates that over the three
years covered by the report, scheme funding valuations showed a significant shift
towards the use of 00 series tables, with a notable number of schemes using the S1
series tables for 2007/2008 valuations.
E.6 The report also showed that half of the 2007/2008 valuations applied either an age
rating or percentage adjustment to the base tables. A year of use approach to
projecting future improvements was used almost exclusively for 2007/2008
valuations.
E.7 Over the three year period covered by the report, there was a shift towards applying
the long cohort adjustment (32% in 2007/08). For 2007/2008 62% of valuations also
applied an underpin to their mortality improvement rates.
61
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Appendix F Response to consultation
F.1 The Government Actuary’s Department’s (GAD’s) public consultation, on the
proposed assumptions to inform my advice on the calculation of contracted-out
rebates for the period 2012 to 2017, closed on 15 November 2010.
F.2 The consultation document put forward possible assumptions for determining the
rebate and invited comments on the proposals. The objective was to collect
constructive evidence based comments to help refine the proposed assumptions.
F.3 Twelve responses to the public consultation were received
46
. The responses were
generally constructive and largely supportive of my proposed approach. The key
points made in response to each of the consultation questions are summarised below,
along with our response. I would like to thank those who replied to the consultation.
F.4 The full public consultation document can be accessed on GAD’s website.
Question 1
Do you agree that we have correctly identified the main relevant changes in the factors
affecting the cost of providing benefits of an actuarial value equivalent to the benefits forgone
by staff who are contracted-out on a defined benefit basis? What other factors do you
consider relevant?
F.5 There was general agreement from respondents that the factors identified were
correct. Some extra suggestions were also made:
One respondent commented that socio-economic factors have become
important in estimating future mortality changes.
One respondent suggested that the introduction of the Pension Protection Fund
and the Employer Debt Regulations are relevant as these have changed the
nature of the pension promise, increasing the extent to which it can now be
considered guaranteed. Although these factors were identified at the last
review, the implications of these are now clearer.
Comment was made that the future existence of many schemes is now likely to
be curtailed, pointing to significantly different financial considerations than those
that would apply to schemes which do not plan to wind-up.
F.6 Overall, there was no consistent clear messages suggesting any change in our view
on the factors.
Question 2
Do you agree that we are planning to advise on a sufficient range of bases? Please suggest,
with reasons, any additional basis that you think we should include. In particular, do you
consider that extra information would be provided by inclusion of an ‘accounting’ basis?
46
Aon Hewitt Associates Limited, the Association of Consulting Actuaries, BBS Consultants and
Actuaries Limited, Buck Consultants Limited, the Confederation of British Industry, First Actuarial LLP,
Hymans Robertson LLP, JLT Benefit Solutions Limited, Mercer Limited, the National Association of
Pension Funds, the Society of Pension Consultants and Towers Watson Limited.
62
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
F.7 Four respondents supported the inclusion of an ‘accounting’ basis, although two
respondents explicitly did not.
F.8 A ‘SCAPE’
47
basis (one respondent) and a ‘buy-out’ basis (two respondents) were
also suggested.
F.9 Many respondents’ replies focused on what the appropriate basis for calculating the
rebate is. Several respondents expressed concern that a rebate calculated on a ‘best
estimate’ basis is not appropriate. Support for a ‘typical funding’ and ‘gilts’ basis was
mixed.
F.10 Comments included:
‘GAD should recognise that contracting out transfers a significant portion of risk
(investment risk, mortality risk, etc.) from the Government to the private sector. In our
view, the ‘best estimates’ option does not sufficiently reward private sector schemes
for taking on this risk. It would be more appropriate for the rebate to be set within the
‘typical funding’ basis to ‘gilts’ basis range.’
‘If it is assumed that the state second pension is a Government-backed promised
entitlement, then a risk free basis of valuation is, in our view, the most appropriate
starting point. Although some downward adjustment to the rebate is determined on
this basis can be justified on the grounds that not all aspects of the S2P are
sacrosanct …, such an adjustment would fall short of a ‘best estimate’ basis…To the
extent that the rebate is below the cost of funding the benefits under the statutory
funding regime, an employer who contracts out is suffering an immediate ‘strain’.’
‘We have concerns over the Government Actuary’s intention to offer a range of
figures to the Secretary of State, particularly such a wide range, as we believe that
this will almost inevitably lead to the Secretary of State choosing a rebate figure
towards the lower end of the range. Contracting back in is not really an option for
many defined benefit schemes …In our view it would be wrong to exploit this position
by basing rebates on anything less than the full economic cost of providing the
benefits concerned.’
‘We are concerned that, partly given the experience at the last review, the
Government will seek to choose a rate no higher than the current rebate of 5.3% and
plump for a best estimate basis… The ‘correct answer’ is to grant a rebate accorded
to the security of the government guaranteeing the alternative SERPS, namely a gilts
basis.’
‘[We] believe that neither a ‘best estimate’ nor a ‘gilts’ basis value would accurately
reflect the cost to employers of contracting out benefits. An estimate calculated on a
typical funding basis, on the other hand, does take into account margins of prudence
not found on a ‘best estimate’ basis which are consistent with the regulatory regime
imposed by the Pensions Act 2004. Thus, the rebate calculated on this basis would
accurately cover the amount which funded DB schemes would typically hold
themselves to cover benefits of equivalent actuarial value to those forgone as a result
of contracting out.’
47
SCAPE, short for Superannuation Contributions Adjusted Allowing for Past Experience, is the
methodology used to set contribution rates for the unfunded public sector pension schemes.
63
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
‘The calculation of a rebate for contracting-out is much closer to a transaction or
pricing exercise than a funding/ budgeting exercise, however, so use of either the
‘best estimate’ or ‘typical funding’ involves an explicit transfer of economic value from
the private sector to the government (and risk in the other direction.)’
F.11 On the specific issue of extra bases, we have not seen any strong arguments for
further bases. In relation to the wider point as to which basis should be selected,
these points should be addressed to the Secretary of State and I do not see them as
an issue for me.
Question 3
Our proposed financial assumptions are influenced by market conditions as at end
March 2010. Given that the rebates calculated will be applied in the future, do you consider it
appropriate for us to update the financial assumptions when we finalise our report?
F.12 The majority of respondents to this question thought that market conditions should be
updated if any changes were significant.
Question 4
Do you agree with the asset allocation underpinning our calculation of the ‘best estimate’
discount rates? If not, what alternative asset allocations would you suggest (please provide
a rationale and evidence)?
F.13 All respondents who supported the inclusion of a ‘best estimate’ basis agreed with the
asset allocation underpinning our calculation of the discount rates.
Question 5
Within the range 2.0% pa to 2.5% pa, what do you consider an appropriate ‘best estimate’
assumption for CPI. Please provide a rationale and evidence to support your view.
F.14 All respondents that commented on the appropriate ‘best estimate’ assumption for
CPI indicated that this should be at the top end of the 2.0% pa to 2.5% pa range, if
not higher. Various reasons for this were cited:
>
>
>
>
Future inflation shocks may be greater on the upside than the downside and
take a number of years to be eliminated. Government policy could change in
future and there is a greater chance that the inflation target would be revised
upwards rather than downwards.
An effective inflation risk premium in the range 1.0% to 1.5% pa, as implied by a
2% CPI assumption, is too high.
If the proposal to include housing costs within an extended CPI is reflected in
the Government’s headline measure of inflation, this would narrow the gap
between the RPI and CPI assumptions.
The Government may not continue to use the CPI measure for inflation in the
long term.
64
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
>
The OBR projections and the conclusions drawn by the Bank of England’s
inflation report are likely to be influenced by the Government target.
F.15 One respondent suggested an RPI assumption of 3.75% pa and a CPI assumption of
3.00% pa.
F.16 I cover my response to this in some depth in section 11.
Question 6
Do you agree with our estimates of the building blocks for deriving the financial assumptions?
If not, please provide alternative assumptions and evidence or a rationale to support these.
F.17 Various criticisms were made about individual assumptions.
F.18 One respondent commented that it is not possible to forecast gilt yields accurately.
F.19 Comments were made that the NAE assumption of 1.5% pa and the gilt yield
assumption of 5.0% pa are at the higher end of the reasonable range. In support of
this, one respondent (a firm of consulting actuaries) cited the range of assumptions
made by its clients’ schemes.
F.20 One respondent made the following points:
Fixed interest gilts are currently below 5.0% pa over the whole range of the
yield curve.
The gap between index-linked and fixed-interest gilt real yields of 1.0% pa is too
high, based on current market conditions.
The corporate bond returns assumption is 0.5% too high based on current
yields of AA bonds.
Even the top end of the equity assumption range is below current market
expectations.
F.21 A couple of respondents did not agree with the method used. One respondent
commented:
‘We find the building block approach as described and applied in this paper somewhat
flawed. For the average assumptions to ‘add up’ and produce an appropriate
average rebate level, there must be a significant number of further assumptions made
and explanations and justifications given about volatilities and degrees of association
between the variables. We suggest that the GAD either uses a fuller model with such
explanations or, uses something more simplified and overtly so.’
F.22 Overall, the balance of responses were mildly supportive but with challenges on
individual assumptions. We see the approach as having deficiencies but consistent
with pensions reviews and with no clear stand-alone alternative. That said, the
approach I have adopted for best estimate does move in the sort of direction implied
in the above.
65
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Question 7
Do you agree that taking a margin of 0.35% pa above the nominal gilt yield is appropriate to
determine the ‘typical funding’ post-retirement discount rate? If not, what alternative method
and / or margin would you suggest (please provide a rationale and evidence)?
F.23 If respondents were happy with our inclusion of a ‘typical funding’ basis then they
were generally in agreement that taking a margin of 0.35% pa above the nominal
yield is appropriate.
Question 8
Do you agree that taking a margin of 1.85% pa above the nominal gilt yield is appropriate to
determine the ‘typical funding’ pre-retirement discount rate? If not, what alternative method
and / or margin would you suggest (please provide a rationale and evidence)?
F.24 If respondents were happy with our inclusion of a ‘typical funding’ basis then they
were generally in agreement that taking a margin of 1.85% pa above the nominal
yield is appropriate.
Question 9
Do you agree that the ‘gilts’ basis post-retirement discount rate should be determined with
reference to the assumed fixed interest gilt return?
F.25 Respondents to this question tended to agree that the discount rate should be
determined with reference to the assumed fixed interest gilt return.
F.26 One respondent considered it strange that index-linked gilts were not used given that
S2P benefits are linked to earnings and inflation.
F.27 One respondent questioned whether we are considering the yield on gilts of an
appropriate duration, given that many schemes are likely to wind-up during the time
period under consideration.
F.28 Overall, there was general agreement with the proposition in the question.
Question 10
For the Best Estimate basis, do you agree with our term-dependent approach to setting the
pre-retirement discount rate? If not, what would you suggest (please provide a rationale and
evidence)?
F.29 There was some agreement amongst respondents to this question that the term-
dependent approach to setting the pre-retirement discount rate is appropriate for the
‘best estimate’ basis.
F.30 One respondent commented that, given many schemes are likely to wind-up in
advance of the 40 year time horizon envisaged, this may not be appropriate.
F.31 Overall, once again, there was general support to the proposition and whilst there
may be some validity in the point raised about winding up, there is little or no general
evidence that this is being taken into account yet for open schemes which do not
intend to cease accrual in the near future.
66
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Question 11
Are you content with the way that the ‘gilts’ basis pre-retirement discount rate is determined
with reference to the fixed interest gilt yield?
F.32 There was general agreement amongst respondents to this question that the discount
rate should be determined with reference to the assumed fixed interest gilt return.
F.33 Just one respondent considered it strange that index-linked gilts were not used given
that S2P benefits are linked to earnings and inflation (although there was no
objection).
Question 12
Do you agree that it is appropriate for the mortality assumptions to be framed in terms of the
suggested SAPS lives tables? If not, please suggest an alternative with a rationale and
evidence.
F.34 The majority of respondents agreed that it is appropriate for the mortality assumptions
to be framed in terms of the suggested SAPS lives tables.
F.35 Comment was made that socio-economic factors have become important factors in
estimating future mortality changes.
F.36 One respondent asked why GAD had not derived its own table.
F.37 One respondent suggested there will be some ‘amounts’ effect, albeit limited by the
cap on band earnings, which suggests that use of the lives table may slightly
understate the cost.
F.38 Overall, we feel that there is sufficient support for our proposals notwithstanding one
or two minor queries.
Question 13
Do you feel that mortality improvements in line with the 2008-based UK principal population
projections are suitable? If not, please suggest an alternative with a rationale and evidence.
F.39 There was general agreement that mortality improvements in line with the 2008-
based UK principal populations are suitable.
Question 14
For the ‘typical funding’ mortality assumption, the margin of prudence is a somewhat arbitrary
assumption and we welcome comments on actual practice and our proposed two year offset.
F.40 Our proposed two year offset was considered reasonable.
F.41 Only one respondent offered any hint of dissent, commenting that this approach does
not distinguish properly between the base table and future improvement assumptions.
They suggested that it would be more transparent to introduce a scaling factor to the
base table or an explicit margin to future improvements.
67
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Question 15
Do you feel that the information provided to the Secretary of State by the ‘gilts’ basis could be
improved if different mortality assumptions (to the ‘typical funding’ basis) were used? If so,
please suggest alternative assumptions with a rationale and evidence.
F.42 There was little support for using a different mortality assumption for the ‘gilts’ basis.
Question 16
Do you consider our proportions married assumptions appropriate for all three bases?
F.43 All respondents to this question agreed that the proportions married assumptions
were appropriate for all three bases.
Question 17
Do you agree with our proposal to omit any remarriage assumption?
F.44 All respondents to this question agreed with our proposal to omit any remarriage
assumption.
Question 18
Are you content for the proposed age difference assumptions to be adopted for all three
bases? If not, please suggest alternatives with a rationale and evidence.
F.45 All respondents to this question were content for the proposed age difference
assumptions to be adopted for all three bases.
Question 19
Do you agree that a 0.2% addition to the rebate is an appropriate allowance for
administrative expenses? If not, please suggest an alternative with evidence.
F.46 The clear majority of respondents considered a 0.2% addition to the rebate to be an
appropriate allowance for administrative expenses.
F.47 Only two respondents also suggested that a small increase in the addition could be
justified.
Question 20
Do you have any comments about the approach to deriving the weights or the weights
themselves?
F.48 As we are focusing on the ‘cost’ in private sector schemes of providing the S2P
benefits forgone, a number of respondents suggested that we exclude public sector
data when calculating the weights.
F.49 One respondent suggested that if it is not possible to exclude public sector data from
the weights, this should be made explicit so that a contingency margin can be
included, if considered appropriate.
68
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
F.50 The points are well made and have been brought into account in the holistic analysis
described in section 11 for best estimate where the point is most critical.
Question 21
When viewed as a whole do you feel that the ‘best estimate’ basis is in line with the definition
set out? If not, please explain, and provide an alternative with evidence and rationale.
F.51 Most respondents were happy with the concept of the ‘best estimate’ basis and how it
is constructed, although many of the criticisms made elsewhere were repeated in this
section.
F.52 Several respondents were clear that while they accept our definitions and analysis
they do not consider it appropriate for the rebate to be calculated on a ‘best estimate’
basis.
F.53 Just one comment was made that the ‘best estimate’ basis will not reflect the actual
investment strategy of many schemes.
F.54 Overall, there were no new points raised here not made elsewhere and probably
majority support.
Question 22
When viewed as a whole do you feel that the ‘typical funding’ basis is in line with the
definition set out? If not, please explain, and provide an alternative with evidence and
rationale.
F.55 Most respondents were happy with the concept of the ‘typical funding’ basis and how
it is constructed, although some of the criticisms made elsewhere were repeated in
this section.
F.56 Some respondents did not consider (as above) it appropriate for the rebate to be
calculated on a ‘typical funding’ basis.
F.57 One comment was made that, due to the scheme specific funding regime, it is not
possible to determine a scheme funding basis that would apply to contracted-out
pension schemes as a whole.
F.58 Overall, there was majority support, in general, for what we suggested.
Question 23
When viewed as a whole do you feel that the ‘gilts’ basis is in line with the definition set out?
If not, please explain, and provide an alternative with evidence and a rationale.
F.59 Most respondents were happy with the concept of the ‘gilts’ basis and how it is
constructed. There were fewer objections to the technicalities or use of this basis.
Question 24
Do you have any comments on our proposed approach to defined contribution schemes?
F.60 There were no objections to our proposed approach to defined contribution schemes.
69
Review of rebates and reduced rates of National Insurance contributions for members of
defined benefit contracted-out pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Question 25
Do you have any other comments on this review of the contracted-out rebates for 2012 to
2017?
F.61 Several respondents made various comments about the different considerations for
private and public sector schemes. (See question 20 above.)
F.62 A number of respondents commented that the smaller the rebate the lower the
incentive to contract out and a low rebate could ultimately lead to closure of defined
benefit pension schemes.
F.63 Comments were made that this review is being carried out against a backdrop of
uncertainty about the future of contracting out and anticipated closure of many
defined benefit pension schemes.
F.64 These comments are potentially valid but outside the scope of my review but, through
publication of this report, will be passed onto the Secretary of State for him to
consider.
Summary of the key responses
F.65 In general, the majority of respondents were broadly content with my proposal to
provide the Secretary of State with rebates calculated on a number of alternative
bases, in order to demonstrate that there is a range of potential approaches and
rebates which might be considered satisfactory. The majority were broadly content
with the assumptions which I had proposed. It should be noted that the best estimate
rate as offered in this report is at the top end of that shown in GAD’s consultation
document (dated 23 August 2010) in part due to the comments received.
70
Review of rebates and reduced rates of National
Insurance contributions for members of contracted-
out defined contribution pension schemes
Report to the Secretary of State for Work and
Pensions by the Government Actuary
Date: 25 January 2011
Author: Trevor Llanwarne
The Right Hon Iain Duncan Smith MP
Secretary of State for Work and Pensions
Department for Work and Pensions
Caxton House
Tothill Street
London
SW1H 9NA
25 January 2011
Dear Secretary of State
Review of rebates and reduced rates of National Insurance contributions for members
of contracted-out defined contribution pension schemes
I am pleased to provide this report setting out my review of the rebates and reduced rates of
National Insurance contributions for members of contracted-out defined contribution pension
schemes for the period from 2012. This is supplementary to my main report, dated
21 January 2011, regarding the rebates for defined benefit schemes.
This report includes recommended rebate rates for defined contribution schemes for
2012/13. It has been produced solely to satisfy the statutory obligations set out in the
Pension Schemes Act 1993. Of course, it is expected that contracting-out on a defined
contribution basis will be abolished from April 2012, so the rebates set out in this report will
not have effect.
I recommend that if the decision to abolish contracting-out on a defined contribution basis is
reversed, then you should seek my further advice on the rebates which may be appropriate
in that situation.
Yours sincerely
Trevor Llanwarne
Government Actuary
Finlaison House 15-17 Furnival Street London EC4A 1AB T: +44 (0)20 7211 2600 W: www.gad.gov.uk
Direct line: +44 (0)20 7211 2620 Fax number: +44 (0)20 7211 2601 Email address: trevor.llanwar[email protected].uk
>
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
1 Overview
1.1 This report sets out my review of the rebates and reduced rates of National Insurance
contributions for members of contracted-out defined contribution pension schemes for
the period from 2012. It is supplementary to my main report dated 21 January 2011,
which covers defined benefit schemes.
1.2 Individuals may contract out through membership of:
a Contracted-Out Salary Related Scheme (COSRS),
a Contracted-Out Money Purchase Scheme (COMPS), or
an Appropriate Personal Pension (APP).
1.3 Those who contract-out through the first arrangement are considered to be ‘contracted-
out on a defined benefit basis’. Those who contract out through the second and third
arrangements are considered to be ‘contracted-out on a defined contribution basis’.
1.4 The Secretary of State is required by statute to lay before Parliament, at intervals of not
more than five years, a report setting out the Government Actuary’s opinion on:
For COSRS, changes since the last report in the factors affecting the cost of
providing benefits of equivalent actuarial value to the Additional Pension that is
forgone by workers who are contracted-out, and
For COMPS and APPs, the percentages required to reflect the cost of providing
benefits of equivalent actuarial value to the Additional Pension that is forgone by
workers who are contracted-out.
1.5 This report covers COMPS and APPs only. My advice in respect of COSRS is provided
in my report dated 21 January 2011.
1.6 Section 15(1) of the Pensions Act 2007 provides that contracting-out on a defined
contribution basis (via Contracted-out Money Purchase Schemes (COMPS) and
Appropriate Personal Pensions (APPs)) will be abolished from a date appointed by the
Secretary of State. The abolition date has not yet been set in legislation, though it is
intended to be 6 April 2012.
1.7 Legal advice is that even if an order is laid confirming the abolition date, the
Government Actuary’s statutory obligation to provide a report recommending rebate
percentages would still be in force. The obligation will remain until the primary
legislation (Pension Schemes Act 1993 sections 42B and 45A) is repealed, which will
not happen before the report for the current review is due.
1.8 Therefore, I must produce a report recommending the percentage rebates for COMPS
and APPs, even though these rebates are not expected to have effect.
75
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
1.9 In Appendix B and Appendix C I provide recommended rebate rates for COMPS and
APPs, for the year 2012/13. They have been produced on the basis that they will not
be used in practice, or at most for the year 2012/13 and then on the basis of no
structural or material changes to the way in which the forgone benefits are determined
in a defined contribution contracting-out environment. Should the decision to abolish
contracting-out on a defined contribution basis be reversed, I recommend that you
should seek my further advice on the rebates which may be appropriate for years
2013/14 onwards and to the extent that there are structural or material changes, I
recommend that you seek my further advice in wider terms.
1.10 Other than the Secretary of State for Work and Pensions, no person or third party is
entitled to place any reliance on the contents of this report and GAD has no liability to
any person or third party for any act or omission taken, either in whole or part, on the
basis of this report.
76
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
2 Introduction to contracting out
2.1 Since the introduction of the state earnings-related pension scheme (SERPS) in 1978,
employees have been able to ‘contract-out’ – that is to give up all or part of their state
Additional Pension benefits. In return they pay lower National Insurance contributions,
or receive rebates of National Insurance contributions paid into their defined
contribution pension schemes or personal pensions.
2.2 The principle of contracting-out continued after the introduction of the state second
pension (S2P), which replaced SERPS in April 2002.
2.3 The legal background of contracting-out is set out in Appendix A. This gives the
relevant statutory references and a summary of the effect on members of contracting-
out.
The contracted-out rebate
2.4 The ‘contracted-out rebate’ refers to the effective reduction of National Insurance
contributions for members of pension schemes, and their employers, or to the rebate
paid to pension schemes, where these schemes are contracted-out of the state
Additional Pension, or to a combination of the two.
2.5 The contracted-out rebate is set having regard to the cost of providing benefits of
equivalent actuarial value to the state Additional Pension that is forgone by workers
who are contracted-out.
COSRS, COMPS and APPs
2.6 Individuals may contract out through membership of:
a Contracted Out Salary Related Scheme (COSRS),
a Contracted Out Money Purchase Scheme (COMPS), or
an Appropriate Personal Pension (APP).
2.7 Those who contract-out through the first arrangement are considered to be ‘contracted-
out on a defined benefit basis’. Those who contract out through the second and third
arrangements are considered to be ‘contracted-out on a defined contribution basis’.
2.8 In respect of members of COSRS, both the employer and employee pay reduced rates
of National Insurance contributions.
2.9 For members of APPs, HM Revenue & Customs (HMRC) makes rebate payments
directly to the pension scheme in respect of contracted-out members.
2.10 A combination of the above occurs for members of COMPS. Reduced rates of National
Insurance contributions are paid and the employer pays the balance of the contracted-
out rebates into the pension scheme. Age related top up payments are also made to
the pension scheme by HMRC.
2.11 A single contracted-out rebate percentage is applied in respect of all members of
COSRS.
2.12 Tables of age related rebates are applied in respect of members of COMPS and APPs.
77
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Abolition of contracting-out on a defined contribution basis
2.13 Section 15(1) of the Pensions Act 2007 provides that contracting-out on a defined
contribution basis (via Contracted-out Money Purchase Schemes (COMPS) and
Appropriate Personal Pensions (APPs)) will be abolished from a date appointed by the
Secretary of State. The abolition date has not yet been set in legislation, though it is
intended to be 6 April 2012.
2.14 Legal advice is that even if an order is laid confirming the abolition date, the
Government Actuary’s statutory obligation to provide a report recommending rebate
percentages would still be in force. The obligation will remain until the primary
legislation (Pension Schemes Act 1993 sections 42B and 45A) is repealed, which will
not happen before the report for the current review is due.
2.15 Therefore, I must produce a report recommending the percentage rebates for COMPS
and APPs, even though these rebates are not expected to have effect.
The Government Actuary’s and Secretary of State’s obligations (COMPS and APPs)
2.16 The Secretary of State is required by statute to lay before Parliament, at intervals of not
more than five years, a report setting out the Government Actuary’s opinion on the
percentages required to reflect the cost of providing benefits of equivalent actuarial
value to the Additional Pension that is forgone by workers who are contracted-out.
2.17 My role, as Government Actuary, is limited to providing advice on the contracted-out
rebates. The power to decide and set the rebate is vested in the Secretary of State for
Work and Pensions.
78
>
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
3 Deriving the defined contribution contracted-out rebates
Calculation methodology
3.1 To derive a contracted-out rebate we have to identify the benefits in the state second
pension scheme which are forgone by contracted-out individuals. Effectively, these are
calculated in the same way as previously under the State Earnings Related Pension
Scheme (SERPS). That is:
a percentage of earnings in a defined band,
payable from state pension age,
revalued in deferment in line with average earnings growth,
increased in payment in line with prices inflation, and
with contingent s
pouse’s benefits of up to 50% of the contributor’s pension.
3.2 The measures of earnings growth and prices inflation are set out in orders.
3.3 I have derived rebates using the assumptions below for a range of ages and for both
sexes. For the recommended rebates, I have selected the higher of each sex’s derived
rebate rate. This is consistent with the 2006 approach.
3.4 Appendix B sets out the scale of age-related rebates for COMPS, including the flat-rate
element. In the past, the flat-rate element of the rebate has been determined as the
lowest rebate applicable for any individual. For the year 2012/13 this is the rebate
applicable for someone aged 15, i.e. 2.4% of upper band earnings.
3.5 The rates shown in Appendix B (COMPS) and Appendix C (APPs) are before the
application of any "cap" on rebates.
3.6 I have been asked by DWP to produce rebate rates which are appropriate for the year
2012/13, solely to meet statutory obligations and noting that the rebates are not
expected to be used in practice given the expected abolition of contracting-out on a
defined contribution basis. If abolition were to be rescinded, I recommend that you seek
my further advice on the rebates which may be appropriate for years 2013/14 onwards
and to the extent that there are structural or material changes, I recommend that you
seek my further advice in wider terms.
Valuation assumptions
3.7 In this context, it is appropriate for me to identify assumptions to derive the rebates by
taking the assumptions used at the last review (in 2006) and updating them to reflect
current economic and demographic conditions. A more rigorous analysis is not
appropriate in the circumstances, and therefore the assumptions adopted here should
not be considered as setting any precedents for any other analysis.
3.8 In order to place a value on the state benefits forgone by contracted-out individuals, we
must adopt assumptions in respect of expected future demographic and financial
experience.
79
>
>
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
3.9 The principal assumptions used for the 2007 – 2012 quinquennium were:
Pre–retirement discount rate (net of
earnings growth)
1.0% - 2.5% pa depending on term to
state pension age
Post–retirement discount rate (net of RPI) 2.0% pa
Pre–retirement mortality None
Post–retirement mortality In line with standard mortality tables and
projections in use at the time
Contingency margin 7.5% of calculated rebates
State pension ages Allowance for increase in SPA ultimately
to age 65 for women.
A full description of the assumptions used is given in Appendix D.
3.10 I have concluded that the assumptions adopted in 2006 are appropriate for deriving the
rebate rates for 2012/13 given the context of this review, with the following exceptions:
Post-retirement discount rate changes from 2% pa net of RPI to 1.75% pa net of
CPI,
The 7.5% contingency margin is reduced to 4%, and
Allowance is made for the increases in state pension ages ultimately to age 68
for the youngest workers (as discussed from section 3.17).
3.11 My rationale for these changes is given below.
3.12 In my opinion, the values adopted for the assumptions in 2006 remain broadly
appropriate in the circumstances of a one year view and I have retained them as far as
is reasonable for this review.
3.13 The pre-retirement discount rates adopted in 2006 represented an estimate of average
long-term investment returns for defined contribution investors adopting a typical
investment strategy, to cover periods of time with a wide mix of market conditions. I
therefore consider these assumptions to remain appropriate for this review.
3.14 The main post-retirement assumptions are:
>
Post-retirement discount rate
Post-retirement mortality, and
A margin.
3.15 In 2006, values were adopted for these parameters which were considered to be
broadly consistent with the retail market price of annuities at the time. These were:
>
>
>
A 2% pa discount rate (net of RPI)
Mortality in line with the standard tables PMA92 / PFA92 with medium cohort
improvements
A margin of 7.5% of calculated rebates
80
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
3.16 It seems reasonable to retain this approach to setting these assumptions at this review
in the circumstances of a one year view. I have tested these assumptions against
recent retail market annuity rates and have concluded that by changing the discount
rate to 1.75% pa (net of CPI), retaining the same mortality and changing the margin to
4% I would obtain annuity values broadly consistent with the pricing of the current
market (noting that CPI-linked annuities are not widely available). I do not need to take
a view on the reasons for this broad consistency, nor to speculate how annuity rates
might change in future, and so I have not done so.
Change of state pension age
3.17 Since the last review, state pension age for younger workers has been increased from
65, ultimately to 68 for the youngest workers. This has two effects:
>
>
Because the state pension (and hence the state pension forgone by contracted-
out workers) is payable later, then it is less valuable in most circumstances.
This is because the payments are further into the future and also payable for a
shorter period, and
Because of the design of the contracted-out deduction
1
, a later state pension
age flows through to a lower rate of accrual.
3.18 Both these features generally lead to a reduction in the appropriate rebate.
3.19 Specimen benefit rates are shown in the table below:
Year beginning Age at
State pension Annual accrual
beginning of
age of benefits *
year
6 April 2012 30
40
50
60
68
67
66
61 (women)
0.385%
0.392%
0.408%
0.588%
60 65 (men) 0.526%
6 April 2016 30 68 0.385%
40
50
60
67
66
65
0.392%
0.400%
0.476%
* based on salary between the Lower Earnings Limit and the Upper Accrual Point
1
The contracted-out deduction is the reduction in Additional Pension applied to individuals who have
accrued pension in a contracted-out arrangement.
81
>
>
>
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Applying the APP Rebate Table
3.20 The state second pension moved to accrual on two bands – 10% and 40% (previously
there were three bands – 10%, 20%, and 40%) from 2010/11. The APP rebates in
Appendix C are expressed as the percentage rebate for the 20% band, even though it
no longer exists.
3.21 In order to calculate the rebate from the table, the percentage in the table should be
doubled for the 40% accrual band and halved for the 10% accrual band.
3.22 The figures have been expressed in terms of the 20% band to enable comparison to
previous reports.
Results
3.23 Appendix B and Appendix C set out recommended rebate rates for COMPS and APPs,
for the year 2012/13. The rates recommended for 2011/12 are also shown for
comparison.
3.24 The rebate rates at each age have fallen. The fall ranges from around 20% at the
youngest ages to 3% at the oldest ages. The reasons for this are set out below:
>
The contingency margin applied has fallen causing the rates to fall. This
affects the rebate rates by the same percentage at all ages.
The net post retirement discount rate used to calculate the rates has fallen,
causing the rebate rates to increase. This affects the rebate rates by broadly
the same percentage at all ages.
As discussed from section 3.17, increases to state pension age result in a
lower value pension and a lower rate of accrual. This causes the rebate rates
at affected ages to fall. Recent changes to state pension age have been
greater at younger ages and so there is a greater reduction in the rates at
these ages.
At older ages the accrual rate falls with year of birth. This reduces the rebates
for the ages effected.
82
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Appendix A Legal background
State second pension and contracting-out
A.1 The state second pension is set out in the Social Security Contributions and Benefits
Act 1992 (SSCBA 1992). The Category A pension (paid to the contributor) is
described from s44. The Category B pension (inherited by the spouse of the
contributor) is described from s48A. The amounts of the state second pension (and
the effect of contracting-out) are set out in Schedule 4A in respect of accrual after
2002/03 and before the ‘flat rate introduction year’. Accrual from the ‘flat rate
introduction year’ is set out in Schedule 4B. The ‘flat rate introduction year’ has not
been prescribed yet, but it is intended that it will be 2012/13.
A.2 The amount of the state second pension forgone by those contracted-out on a defined
benefit basis is defined identically in both Schedule 4A and 4B, so no difficulty arises
in this regard from the uncertainty around the definition of the ‘flat rate introduction
year’.
A.3 For those contracted-out on a defined contribution basis, only Schedule 4A provides
definitions of contracted-out benefits, since the ‘flat rate introduction year’ is expected
to be synchronised with the abolition of contracting-out on a defined contribution
basis.
National Insurance rebates
A.4 The Pension Schemes Act 1993 (PSA 1993), sections 40 to 49 inclusive, makes
provision for members of pension schemes who are contracted-out of the state
second pension (and the sponsoring employers of those schemes) to pay reduced
rates of National Insurance contributions and/or to have their pension schemes
receive corresponding payments from HM Revenue & Customs (HMRC).
A.5 PSA 1993 s42 requires a review to be carried out by the Government Actuary, at
intervals of not more than five years, of the contracted-out rebate for those
contracted-out on a defined benefit basis (members of COSRS). The review should
report on any changes in the factors which, in the Government Actuary’s opinion,
affect the cost of providing benefits of an actuarial value equivalent to that of the state
pension benefits which are given up by or in respect of members of these schemes.
The legislation does not require the Government Actuary to recommend percentage
rebates, though in practice the Department for Work and Pensions commissions GAD
to provide advice on the level of rebate values (and there is no legislative bar
preventing this).
A.6 Separate requirements exist for members of Contracted-Out Money Purchase
Schemes (COMPS) under PSA 1993 s42B. COMPS are occupational defined
contribution schemes. Contracted-out rebates are age-related for members of
COMPS. The Government Actuary must report on the percentage age-related
rebates which are required in his opinion to reflect the cost of providing benefits of an
actuarial value equivalent to that of the state pension benefits which are given up by
or in respect of members of COMPS.
83
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
A.7 PSA 1993 s45A specifies age related rebates for Appropriate Personal Pensions
(APPs). These rebates are paid by HMRC direct to the schemes. The Government
Actuary must report on the age-related rebates which are required in his opinion to
reflect the cost of providing benefits of an actuarial value equivalent to that of the
state pension benefits which are given up by or in respect of members of APPs.
A.8 The same sections of PSA 1993 require any order by the Secretary of State for Work
and Pensions to vary the rate of reduction in National Insurance contributions or
rebates, to which the Government Actuary’s reports would relate, to be made at least
one complete tax year before they come into force and to be laid within 5 years of the
previous reports and draft order. Hence for new rebates to apply from 6 April 2012,
the relevant orders and the Government Actuary’s reports must be laid before
Parliament by 28 February 2011, since the previous order was laid on 1 March 2006.
Changes since the previous review
A.9 For people in contracted-out employment, section 10(3)(b) of the Pensions Act 2007
(PA 2007) provides that the state second pension moves to accrual on two bands
(previously there were three bands) from 2010/11.
A.10 Section 15(1) of PA 2007 provides that contracting-out on a defined contribution basis
(via COMPS and APPs) will be abolished from a date appointed by the Secretary of
State. The abolition date has not yet been set, but is expected to be April 2012, at the
same time as the Band 1 accrual of the state second pension moves to a flat rate.
A.11 Section 13 of PA 2007 provides that state pension age will rise from 65 to 68 over the
period 2024 to 2046.
A.12 At the June 2010 budget, the Chancellor of the Exchequer announced that state
second pension will be increased in payment in line with the Consumer Prices Index
rather than the Retail Prices Index.
A.13 As part of the 2010 spending review, the Government announced plans to accelerate
the rise in state pension ages. Subject to Parliamentary approval the state pension
age should rise to age 65 for all by 2018 and age 66 by 2020.
Reduced state pension
A.14 The effect of contracting-out on a member’s state second pension is set out in the
Social Security Contributions and Benefits Act 1992 (SSCBA 1992) Schedules 4A
and 4B.
A.15 For those contracted-out on a defined benefit basis, members give up a percentage of
their earnings between the Qualifying Earnings Factor (that is, the Lower Earnings
Limit expressed as an annual amount) and the Upper Accrual Point (expressed as an
annual amount).
A.16 The percentage is 20 divided by the number of ‘relevant years’ in the member’s
‘working life’. ‘Working life’ (defined in SSCBA 1992 schedule 3 paragraph 5(8)) is the
period between the tax year the member attains the age of 16, and the tax year
immediately before attaining state pension age. ‘Relevant years’ (defined in SSCBA
1992 s44(7)) are restricted to 1978/79 and later.
A.17 The benefits forgone would otherwise have been payable at state pension age.
84
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
A.18 From the ‘flat rate introduction year’ the state second pension itself is amended. For a
few members (older and higher earning) the state second pension is reduced to zero.
That is, the specified deduction for contracted-out workers exceeds their state second
pension.
Inherited benefits (simplified summary)
A.19 Where both the contributor and the spouse (including civil partners) are over state
pension age at the date of the contributor’s death then the spouse generally becomes
entitled to a pension equal to half of the contributor’s accrued state second pension
(though in some cases inheritance is capped at a maximum value).
A.20 If the spouse is under state pension age then the inherited pension is not paid until
state pension age unless there are dependent children. If the spouse is between 45
and 55 and there are no dependent children then the inherited pension is reduced, as
well as being delayed until state pension age. If the spouse is under 45 and there are
no dependent children then the inherited pension is eliminated.
Reduced National Insurance Contributions
A.21 The Pension Schemes Act 1993 s41(1), (1A) and (1B) provide that the Class 1
National Insurance contributions payable in respect of the member contracted-out on
a defined benefit basis should be reduced by a percentage of their earnings between
the Lower Earnings Limit and the Upper Accrual Point.
Defined contribution schemes
A.22 For the time being, COMPS have the same state pension reduction as COSRS.
Members of APPs generally forgo their entire entitlement to state second pension.
A.23 The National Insurance rebates for those contracted-out on a defined contribution
basis are age related, and may differ between COMPS and APPs.
85
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Appendix B Contracted-out rebates – COMPS
Rebates proposed for members of contracted-out money purchase pension schemes
Percentage of upper band earnings, with no cap
Age last birthday on last day of Tax year
preceding tax year 2011-12 Current 2012-13
Proposed
15 3.0% 2.4%
16 3.0% 2.5%
17 3.1% 2.5%
18 3.2% 2.6%
19 3.3% 2.7%
20 3.4% 2.7%
21 3.4% 2.8%
22 3.5% 2.9%
23 3.6% 2.9%
24 3.7% 3.0%
25 3.8% 3.1%
26 3.9% 3.1%
27 4.0% 3.2%
28 4.1% 3.3%
29 4.2% 3.4%
30 4.3% 3.4%
31 4.4% 3.5%
32 4.5% 3.6%
33 4.6% 3.7%
34 4.7% 4.1%
35 4.8% 4.2%
36 5.0% 4.3%
37 5.2% 4.4%
38 5.3% 4.5%
39 5.5% 4.7%
40 5.6% 4.8%
41 5.8% 5.0%
42 6.0% 5.1%
43 6.1% 5.7%
44 6.3% 5.9%
45 6.5% 6.0%
46 6.6% 6.2%
47 6.8% 6.3%
48 7.0% 6.5%
49 7.3% 6.6%
50 7.6% 6.9%
51 8.0% 7.3%
52 8.4% 8.2%
53 8.8% 8.6%
54 9.2% 9.0%
55 9.6% 9.4%
56 10.5% 9.8%
57 11.6% 10.8%
58 12.9% 12.0%
59 14.3% 13.3%
60 11.8% 11.5%
61 12.2% 11.9%
62 12.7% 12.4%
63 13.2% 12.8%
86
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Appendix C Contracted-out rebates – APPs
Contracted-out rebates proposed for members of appropriate personal pension
schemes
Percentage of earnings applicable to band 3 (“20% accrual rate band”), with no cap
Age last birthday on last day of Tax year
preceding tax year 2011-12 Current 2012-13 Proposed
15 4.7% 3.8%
16 4.7% 3.9%
17 4.8% 4.0%
18 4.9% 4.0%
19 5.0% 4.1%
20 5.0% 4.1%
21 5.1% 4.2%
22 5.2% 4.3%
23 5.2% 4.3%
24 5.3% 4.4%
25 5.4% 4.5%
26 5.5% 4.5%
27 5.6% 4.6%
28 5.6% 4.6%
29 5.7% 4.7%
30 5.8% 4.8%
31 5.9% 4.8%
32 6.0% 4.9%
33 6.1% 5.0%
34 6.1% 5.4%
35 6.3% 5.5%
36 6.4% 5.6%
37 6.6% 5.7%
38 6.7% 5.8%
39 6.9% 5.9%
40 7.0% 6.1%
41 7.1% 6.2%
42 7.3% 6.3%
43 7.4% 6.9%
44 7.5% 7.0%
45 7.6% 7.1%
46 7.8% 7.3%
47 7.9% 7.4%
48 8.0% 7.5%
49 8.3% 7.6%
50 8.6% 7.9%
51 8.9% 8.2%
52 9.3% 9.1%
53 9.6% 9.4%
54 10.0% 9.7%
55 10.3% 10.0%
56 11.1% 10.4%
57 12.1% 11.3%
58 13.1% 12.3%
59 14.3% 13.4%
60 12.1% 11.7%
61 12.4% 12.1%
62 12.8% 12.4%
63 13.1% 12.7%
The percentage in the table should be doubled for the “40% accrual band” and halved for the “10% accrual band”.
87
Review of rebates and reduced rates of National Insurance contributions for members of
contracted-out defined contribution pension schemes
Report to the Secretary of State for Work and Pensions by the Government Actuary
Appendix D Assumptions adopted at the previous review
COMPS APPs
Financial
Pre-retirement discount rate
(net of earnings growth)
1% pa just before SPA
2% pa 15 yrs to SPA
2.5% pa >30 yrs to SPA
1% pa just before SPA
2% pa 15 yrs to SPA
2.5% pa >30 yrs to SPA
Post-retirement discount rate
(net of pension increases)
2% pa 2% pa
Gross pre-retirement discount
rate
5% pa 5% pa
Demographic
Pre-retirement mortality None None
Post-retirement mortality PMA92 / PFA 92
medium cohort
PMA92 / PFA 92
medium cohort
Proportions ‘married’ 100% at SPA 100% at SPA
‘Marital’ age differences In line with the 2001
national census data
In line with the 2001
national census data
Administration expenses 0.2% of band earnings 1% a year pre
retirement (reduces
discount rate)
Contingency margin 7.5% of calculated
rebate
7.5% of calculated
rebate
Rationale for contingency
margin
To cover loadings in
annuity prices
To cover loadings in
annuity prices
State Pension Ages Allowance for increase
in SPA ultimately to
age 65 for women
Allowance for increase
in SPA ultimately to
age 65 for women
88
Report by Secretary of State for Work and Pensions in accordance with
sections 42(1)(b), 42B(1)(b) and 45A(1)(b) of the Pension Schemes Act 1993
Report by Secretary of State for Work and Pensions in accordance with
sections 42(1)(b), 42B(1)(b) and 45A(1)(b) of the Pension Schemes Act 1993
1. In accordance with the Pension Schemes Act 1993, I am laying before
Parliament my report on the reduced rates and rebates of National Insurance
contributions applicable in respect of members of contracted-out occupational
pension schemes and appropriate personal pension schemes.
2. This report, the accompanying draft Order and the Government Actuary’s
report address the percentage reduction in National Insurance contributions
applicable to employers and employees of salary related contracted-out schemes
(defined benefit schemes) and money purchase schemes (defined contribution
schemes); and the rebates applicable to members of money purchase schemes
and approved personal pension schemes (APPs).
3. However, it is planned to abolish contracting-out on a defined
contributions basis on 6 April 2012. When the enabling legislation for abolition
(section 15(1) of the Pensions Act 2007 and provisions in primary and secondary
legislation which make consequential amendments) is brought into force,
sections 42B and 45A of the Pension Schemes Act 1993 will be repealed.
However, until that date, my responsibility under sections 42B and 45A of the
Pension Schemes Act 1993 that defined contribution rebates are reviewed at
least every five years remains. But because these figures are not expected to be
used, they are only provided for one tax year.
4. This report will therefore focus primarily on the reduction in National
Insurance applicable to those contracted out on a defined benefit basis. That
reduction is expected to reflect the cost to schemes of providing benefits of
equivalent actuarial value to the State Pension that is foregone by employees
who are contracted out.
(1) Pension Schemes Act 1993 section 42(1)(b)
Reduced rate of Class 1 National Insurance Contributions for members
of salary-related contracted-out schemes
5. The first part of my report, required by Section 42(1)(b) of the Pension
Schemes Act 1993 concerns the level of reduction in National Insurance for
employers and employees within salary-related contracted out occupational
pension schemes. Currently Section 41(1A) and (1B) of the Pension Schemes
Act 1993 (which incorporates rebates from the last Rebate Order
1
), provides for a
reduction in the combined employer and employee contributions by a total of
5.3% of relevant earnings (split as 1.6% primary (employee) and 3.7% secondary
(employer).
1
The Social Security (Reduced Rates of Class 1 Contributions, Rebates and Minimum
Contributions) Order 2006 (SI2006/1009)
91
6. Since the last report the definition of relevant earnings has changed and
now refers to earnings between the lower earnings level and the upper accrual
point. This is a point of detail, however, rather than a material fact which affects
the Rebate.
7. In a departure from previous reports, the Government Actuary has
provided three alternative approaches for valuing the cost of the benefits
foregone by contracted-out workers in salary-related schemes: a “best estimate”
basis, a “typical funding” basis, and a “gilts” basis.
8. I welcome the advice provided by the Government Actuary which has
informed my decisions on the level of the rebate. The reduction in National
Insurance contributions is provided at considerable current cost to the tax
payer. The Government has a duty, therefore, to ensure that the rebates are
set at a level which is fair to all. In view of those considerations, I believe that
adopting the best estimate basis is the correct approach.
9. In calculating the best estimate rate which should apply to earners in
salary-related schemes, I have considered the changes affecting the cost of
providing benefits of equivalent actuarial value to the SERPS foregone which the
Government Actuary has detailed in his report, for example, the assumptions
made concerning changes in membership details, increased longevity, the future
change in indexation of the State Additional Pension by the Consumer Price
Index, rather than the Retail Price Index, and current economic conditions. In my
view, the assumptions made by the Government Actuary are justifiable.
10. The Government Actuary has also helpfully provided two rebate rates for
each approach; one which takes into account the existing arrangements for State
Pension Age, and one which takes into account the proposed changes the
Government will be legislating for which will see State Pension age rise to age 66
from April 2020. I have decided that the proposed rise in State Pension age to
66 needs to be reflected in the revised rebate rate. I have therefore decided, in
line with the Government Actuary’s best estimate approach, to decrease the
reduction in the rate of National Insurance contributions to 4.8% of relevant
earnings. If, for any reason, the proposed changes to State Pension age are not
implemented, I will consider the need to conduct a further review of rebate rates
before the end of a further 5 year period – the time by which I must review rebate
rates again.
11. I propose broadly to maintain the division of the reduction in the level of
primary and secondary contributions. This will therefore mean that the level in the
reduction of primary contributions will be reduced to 1.4% and the reduction in
the secondary contribution will be reduced to 3.4%.
92
(2) Pension Schemes Act 1993 section 42B(1)(b)
Determination and alteration of reduced rates of contributions and
rebates for members of money purchase contracted-out schemes
applicable under Section 42A of that Act
12. The second part of my report, required by Section 42B(1)(b) of the
Pension Schemes Act 1993 deals with the appropriate percentages of the
relevant earnings which need to be specified for members of money purchase
contracted-out schemes.
13. The Government Actuary’s report sets out, by reference to the age of the
individual, the percentage of relevant earnings which he considers reflects the
cost of providing benefits of an actuarial value equivalent to the State Second
Pension foregone. The percentage is delivered in two parts – a flat-rate element
(which takes the form of a reduction in Class 1 National Insurance contributions)
and an age-related rebate paid to the scheme after the end of the tax year to
which it relates.
14. This part of the report is, however, provided simply to meet an existing
legal requirement which will be repealed before these proposed rates have
effect, on account of the proposed abolition of contracting-out on a defined
contributions basis on 6 April 2012. It is not expected that these rates should be
used. Hence the Government Actuary has provided rebate rates for one year
only.
15. With this in mind, I see no reason to alter the rates provided by the
Government Actuary in his report. I have therefore decided to decrease the flat
rate element for contracted out money purchase schemes to 2.4% of relevant
earnings for members of contracted-out money purchase schemes
.
I propose to
split the flat-rate rebate between the primary contribution at 1.4% and the
secondary contribution at 1.0%.
16.
I have also decided to decrease the level of the age-related element of
the rebate as the Government Actuary recommends. I have also concluded that
the rebates should continue to be subject to a cap. The current level of this cap is
7.4% of relevant earnings. I have decided to maintain this at 7.4%.
(3) Pension Schemes Act 1993 section 45A(1)(b)
Determination and alteration of rates of minimum contributions for
those with appropriate personal pensions applicable under Section 45
of that Act
17. The third part of my report, required by section 45A(1)(b) of the Pension
Schemes Act 1993 deals with the age-related percentages of the relevant
earnings which need to be specified for members of appropriate personal
pension schemes.
93
18. As for members of contracted-out money purchase schemes, the
Government Actuary’s report sets out, by reference to the age of the individual,
the percentage of relevant earnings which he considers reflects the cost of
providing benefits of an actuarial value equivalent to the State Second Pension
foregone.
19. For the reasons I have already given concerning the abolition of
contracting-out on a defined contribution basis on 6 April 2012, I, again, see no
reason to alter the rates provided by the Government Actuary in his report. I have
therefore decided to
decrease the age-related rebates for holders of appropriate
personal pensions in line with the Government Actuary’s report.
20. As with contracted-out money purchase schemes, these age-related
rebates are subject to a cap, currently set at 7.4%. I have decided to maintain the
cap for appropriate personal pensions and set it also at 7.4%. The figures (and
therefore the cap) in Schedule 2 of the Order take account of the change from
the 2010/11 tax year to two accrual bands of 10% and 40%, but are based on the
20% accrual band provided in the Government Actuary’s report.
(4) Revaluation of Guaranteed Minimum Pensions (GMPs)
21. The final part of the Government Actuary’s defined benefit report
recommends the percentage by which GMPs should be revalued for contracted-
out members who leave pensionable service in the period 2012 to 2017 where
the scheme has adopted the fixed rate revaluation method (currently 4.0% a year
compound for those leaving since April 2007). The Government Actuary has
proposed a rate of 4.75% compound based on national average earnings growth
of 4.25% with the addition of 0.5%. The addition of 0.5% is to compensate for the
conversion of a variable rate of revaluation (linked to actual average earnings
growth) to a fixed rate for which there has to be a price.
22. I welcome the Government Actuary’s assumptions. We will consider
whether to make a further change to the existing regulations later in the year.
94
1VCMJTIFECZ5405IF4UBUJPOFSZ0GGJDFBOEBWBJMBCMFGSPN
0OMJOF
XXXUTPTIPQDPVL
.BJM5FMFQIPOF'BY&NBJM
540
10#PY/PSXJDI/3(/
5FMFQIPOFPSEFST(FOFSBMFORVJSJFT
0SEFSUISPVHIUIF1BSMJBNFOUBSZ)PUMJOF-P$BMM
'BYPSEFST
&NBJMDVTUPNFSTFSWJDFT!UTPDPVL
5FYUQIPOF
5IF1BSMJBNFOUBSZ#PPLTIPQ
#SJEHF4USFFU1BSMJBNFOU4RVBSF
-POEPO48"+9
5FMFQIPOFPSEFST(FOFSBMFORVJSJFT
'BYPSEFST
&NBJMCPPLTIPQ!QBSMJBNFOUVL
*OUFSOFUIUUQXXXCPPLTIPQQBSMJBNFOUVL
540!#MBDLXFMMBOEPUIFS"DDSFEJUFE"HFOUT
$VTUPNFSTDBOBMTPPSEFSQVCMJDBUJPOTGSPN
540*SFMBOE
"SUIVS4USFFU#FMGBTU#5(%
5FM 'BY
