Memorandum submitted by the Association
of British Insurers (ABI)
EXECUTIVE SUMMARY
1. The ABI wholeheartedly supports the Pensions
Commission's view that pension costs can be substantially reduced
and saving increased, via:
— compulsory employer contributions;
— reduced regulatory costs; and
— reforms to means-testing.
2. However, we do not agree with the Pensions
Commission that it is necessary to set up an entirely new, state-sponsored,
pension organisation. Such an approach would simply expose the
taxpayer to high set-up costs and the financial risks associated
with major Government IT projects.
3. Instead, the Government should make use
of the insurance industry's proven experience and IT infrastructure.
This approach—Partnership Pensions—would:
— deliver the Pensions Commission's
goal of low-cost "personal account" pensions by using
auto-enrolment and compulsory employer contributions;
— harness competition to increase participation
rates and saving levels, improve service and exert downward pressure
on charges;
— offer choice, for those who want
it, and a good "no choice" default option for those
who do not;
— protect good existing workplace pensions
and the broader pensions market;
— operate at a charge level at, or
only slightly higher, than those actually likely under the Pensions
Commission's proposals; and,
— be up and running far more quickly
than a state-sponsored scheme.
4. Charge levels under any model will depend
on a range of factors: contribution levels; persistency; the costs
of collecting contributions; administration; fund management;
and regulation. But charges as low as 0.3% are not possible under
the Commission's model. Deloitte's analysis for the ABI shows
that the National Pensions Savings Scheme (NPSS) proposed by the
Commission would require an Annual Management Charge (AMC) of
at least 0.45% and possibly as high as 0.6% of fund value. Moreover,
the Swedish Premium Pension Authority (PPM), often cited as an
example of how to achieve low charges, currently shares the costs
of collection with the state PAYG pension. The PPM confirms that,
without this, the average charge would be over 0.8% AMC and the
cheapest charge would be more than 0.5% AMC.
5. To ensure value for money and to address
the need for robust governance of the new scheme, we also propose
the creation of a new market-focussed economic regulator, the
Retirement Income Commission (RIC). This body would:
— make economic market assessments
of costs, charges and contributions and recommend a charge cap
if necessary;
— monitor savings behaviour and the
effectiveness of policy, in particular to ensure that "levelling
down" of existing provision does not occur;
— carry out certain tasks necessary
for the new system to work; and
6. Our proposals for Partnership Pensions
and their regulation draw on the lessons of Stakeholder pensions:
— the policy and regulatory framework
needs to be designed as an integrated whole—for Stakeholder
pensions, the sales regulation makes it difficult to advise and
persuade savers to join a pension within the price cap;
— employer contributions and reduced
reliance on means-tested pension benefits are needed to ensure
saving in the new scheme is "suitable" for almost all
savers, so avoiding the need for costly regulation—the absence
of an employer contribution and increased use of means-tested
benefits preventes Stakeholder pensions reaching their full potential;
— a charge cap should be considered
only if there has been a demonstrable market failure over a period
of time, and after a sound, transparent and extensive analysis
by an economic regulator (the RIC)—the absence of proper
economic regulatory analysis and an opaque process makes the Stakeholder
pension price cap unfit for purpose;
— the regulator should start from the
assumption that saving is in the interest of the scheme member,
rather than placing the main emphasis on seeking to prevent "unsuitable"
sales—the Basic Advice regime for Stakeholder pensions has
the emphasis the other way round, so inhibiting saving.
7. Whatever decision the Government takes
on a new model for private pension saving, action could be taken
very quickly to address the weaknesses within the current Stakeholder
pension framework. This would mark a useful first step in helping
more people save for retirement.
1. INTRODUCTION
1.1 The ABI welcomes the Treasury Select
Committee's inquiry into the design of the proposed National Pension
Saving Scheme (NPSS) and the role of financial services regulation.
The Pensions Commission's Second Report sets out a compelling
vision of increased saving through auto-enrolment and compulsory
employer contributions. If we learn the lessons of Stakeholder
pensions and overhaul the regulatory framework accordingly, a
major opportunity can be seized for the country and for millions
of ordinary savers.
1.2 This submission sets out:
— our agreement with the Pensions Commission's
vision and the strengths and weakness of its proposed implementation
strategy, and how the ABI's proposals for Partnership Pensions
would provide a new beginning for private pension saving (see
Partnership Pensions: A New Model for Retirement Saving,
February 2006);
— the main determinants of costs and
charges, and the likely level of charges under the Pensions Commission's
proposals;
— the design of a new Retirement Income
Commission; and
— lessons from the Stakeholder experience.
1.3 The ABI represents the interests of
the UK's insurance industry. The Association has around 400 member
companies who provide over 97% of insurance business in the UK.
The UK insurance industry pays out £156 million every day
in pensions and life insurance benefits.
2. A NEW BEGINNING
FOR PRIVATE
PENSION SAVING:
PARTNERSHIP PENSIONS
2.1 The Pensions Commission proposes a radical
change in the current model of private pensions in order to reach
new savers with average and low incomes. To achieve this, the
Commission proposes to cut costs in a number of ways:
— auto-enrolment, which would automatically
place employees in a pension scheme, so removing the cost to the
pension company of seeking new business and of persuading staff
to join occupational schemes;
— mandatory employer contributions
at 3% and reduced reliance on means-testing, to ensure all pension
saving is "suitable" and can be undertaken without regulated
advice, so cutting regulatory costs.
2.2 The ABI unequivocally supports these
aims and believes that, together, they hold the key to a successful
new beginning for pension saving. However, we do not agree with
the Pensions Commission that it is necessary to set up an entirely
new, state-sponsored, pension organisation. We believe that such
an approach would simply expose the taxpayer to high set-up costs,
uncertain future liabilities, and the financial risks associated
with major Government IT projects.
2.3 We believe that the Government would
be better advised to build on the insurance industry's existing
systems and expertise. The ABI's proposals, "Partnership
Pensions", developed at the request of the Minister for Pensions
Reform, show how this can be done.
2.4 Partnership Pensions would deliver the
Pensions Commission's goal of low-cost "personal account"
pensions for all. They would achieve this by using auto-enrolment
and compulsory employer contributions in the context of a lighter,
more proportionate, regulatory regime and reduced reliance on
means-testing. But, by using the insurance industry's proven skills,
know-how and IT infrastructure, they would avoid the risks to
the taxpayer and Government of establishing an entirely new organisation.
2.5 Our proposals would harness competition
to increase participation rates and saving levels, improve service
and exert downward pressure on charges. They would offer choice—of
provider, fund, and service level—to those who want it,
and a good "no choice" default option for those who
do not. Under Partnership Pensions, if an individual wished to
shop around for a provider who offers additional information and
help with saving, they could do so. If they did not want to exercise
choice, they would be automatically enrolled into a Partnership
Pension designated by their employer or selected automatically
(from a "carousel""). They would then be placed
in a default lifestyle fund. They would remain with this provider,
even if they changed employer, unless they actively chose to switch
to a different pension or were offered a better pension.
2.6 Similarly, companies would not have
to choose a provider if they did not wish to do so, as they could
be allocated one via a carousel. Indeed, Partnership Pensions
are specifically designed to place as small a demand as possible
on employers—for many, the task would be no harder than
sending a pension contribution to the employee's unique pension
account.
2.7 Partnership Pensions would operate at
a charge—probably in a range between 0.6% to 0.75% AMC—the
same, or only a little higher, than the NPSS in practice. (The
independent analysis and international comparisons set out in
chapter 3 suggest that the NPSS would be likely to charge at least
0.45% AMC, and quite possibly 0.6% AMC or more.)
2.8 To ensure value for money and address
the need for robust governance of the new scheme, we also propose
the creation of a new market-focused economic regulator, the Retirement
Income Commission. This would: make economic assessments of costs,
charges and contributions; monitor savings behaviour and the effectiveness
of policy; carry out tasks necessary for the successful operation
of the new system; and encourage saving. Its role is discussed
further in the annex.
2.9 Moreover, as insurers already have the
skills and infrastructure necessary to establish Partnership Pensions,
the system could be up and running far more quickly than a state-sponsored
scheme. Significantly lower set-up costs would make Partnership
Pensions cheaper than the Pensions Commission's proposal for at
least the first decade of operation.
3. THE COSTS
AND CHARGES
OF A
NEW PENSIONS
MODEL
The Determinants of Costs and Charges
3.1 The operational cost, and charges, of
any new pensions model will depend on:
— The cost of collection, which
will fall as a % of fund value as contribution levels rise and
vice-versa. Costs will be lower if existing collection infrastructure
is used, such as BACs and the IT systems already in place for
Stakeholder pensions, rather than designing and building an entirely
new system.
— The cost of policy administration,
which will be lower if automation is used instead of manual
processing. (In addition, if the Government invited providers
to tender for bundled administration services, this too would
have a bearing on costs—a short contract length to aid performance
management would necessarily involve higher costs.)
— The cost of capital, both
working capital which will be similar in all models, and regulatory
capital which might be higher for private sector providers, is
also an important factor.
— The cost of fund management, which
will depend on how actively the fund is managed, the size of the
fund, and the market power of the service purchaser.
— Whether or not persuasion is necessary
to encourage saving. Entirely voluntary saving requires persuasion,
which is costly. Mandatory saving is cheaper. Automatic enrolment
falls somewhere between the two.
— The number of pensions per person.
Costs would be lower if there is only one pension per lifetime,
because the set-up costs would be spread over a full working life.
Costs would be higher if individuals have multiple pensions.
— Incentives likely to increase
or decrease costs. Without efficiency pressures or incentives
to reduce costs, costs will stay level or rise.
— The cost of regulation. Costs
will be determined by the regulatory framework. The more the system
can be designed to avoid the need for costly sales regulation,
the lower costs will be.
Charges: why 0.3% is more than optimistic
3.2 The Pensions Commission argues that
their model for the NPSS would involve an annual charge of 0.3%
of an individual's total pension fund (0.3% AMC). Unsurprisingly,
commentators have found this an attractive prospect. But the Commission
itself recognises in its Second Report that the figure is intended
to be aspirational, rather than a cap to be implemented quickly.
Even so, more broadly-based modelling than that used by the Commission,
and international comparisons, both suggest that no new pension
system—whether the Commission's NPSS proposals or the ABI's
Partnership Pensions—would in practice be likely to operate
at such a low level of charge.
3.3 All charge estimates are based on modelling
of pension systems which are very different to the current one.
They are, therefore, inherently uncertain. Small changes in assumptions
make very big differences. The ABI asked Deloitte to undertake
a detailed analysis of the costs and charges likely to arise under
the Pensions Commission's proposal for the NPSS. Using the Commission's
base assumptions, Deloitte found that charges were likely to be
around 0.45% AMC rather than 0.3%. Costs had been underestimated
in three main ways:
— policy set-up and fund management
costs would be likely to be higher than assumed by the Commission
(adding 0.06%);
— the NPSS is aimed at all those in
the target population, not just the median earner used in the
Pensions Commission's modelling (adding 0.04%); and
— premiums and charges would be incurred
on a monthly, not an annual basis as assumed by the Pensions Commission
(adding 0.05%).
3.4 We note that the Investment Management
Association (IMA), which has broadly endorsed the Commission's
model, has estimated the likely charge as closer to 0.5% than
0.3% AMC. Moreover, our own modelling shows that assumptions slightly
different from those used by the Commission, but highly plausible,
would increase the charge to around 0.6% AMC:
— if the improvement in persistency
(ie how long people keep paying into the same pension) was only
half the level assumed by the Pensions Commission, the charge
would rise by about 0.11%. Increasing labour mobility makes this
a reasonable assumption;
— if automatic enrolment only achieved
65% participation rather than the 80% assumed by the Commission,
charges would rise by 0.02%. (80% participation rates are only
seen when the employer actively encourages enrolment. Commercial
pressures might well mean that smaller UK employers would be less
encouraging. And uneven contributions from lower earners seem
highly probable.)
3.5 These uncertainties apply to other models
too. The National Association of Pension Funds (NAPF) estimated
that the Commission's model would cost 0.45% AMC, and proposed
their own model of Super Trusts at charges a little over 0.4%
AMC. However, according to the NAPF's own assessment, if plausible
alternative assumptions are used, the charge for Super Trusts
would increase to around 0.7% AMC.
International Experience
3.6 International experience has been cited
in support of charges as low as 0.3% AMC. Close examination of
that experience shows that charges at this level have not, in
practice, been achieved. Application of international experience
to the UK needs to be done carefully. For example, Sweden, which
operates a system very like the Commission's model, currently
has an average charge of over 0.6% AMC. The cheapest fund, the
large default fund, charges a little below 0.4% AMC. But the cost
of administering the Swedish system is shared with the state PAYG
pension. The Swedish authorities have confirmed to the ABI that
if this were not the case, charges could be around 0.2% AMC higher,
so the average charge would be over 0.8% AMC and even the very
cheapest fund would have to charge over 0.5% AMC.
3.7 Several commentators have also suggested
that the USA and New Zealand offer indications of how much an
NPSS might charge. However, in neither case is such a system in
operation so it is too early to draw any reliable conclusions
about the charges that would arise.
Other Factors
3.8 Our analysis is not a "worst case"
scenario. We have not, for example, factored into the NPSS costs
a repeat of the Swedish experience, which saw the costs of two
failed attempts to build IT infrastructure written off by the
Swedish Government.
3.9 Nor is it a dynamic analysis. We have
not attempted to factor in how charges would fall over time. But
the ABI model preserves more choice for both employers and employees
and therefore incentivises both price and service competition
more effectively than the NPSS. Such savings can be significant:
between 1990 and 2004, management expenses including commission
for long term insurance business fell from 2.6% to 1%. (1.6% to
0.6% if commission costs are excluded.) One-off competitions for
lengthy fund and administration mandates are likely to deliver
less benefit than this more dynamic market competition, and lead
to less choice and flexibility in the event of supplier failure.
Final Thoughts on Charges
3.10 Both international comparisons and
modelling show that the charge for the Pension Commission's model
would be around 0.5% AMC rather than 0.3% AMC. There is therefore
only a small difference compared with the likely charge for Partnership
Pensions (between 0.6% and 0.75% AMC). Partnership Pensions would
also provide other benefits and avoid other risks, as outlined
elsewhere in this paper.
3.11 That said, the decision on which model
to use should not be based solely on the likely charge. We believe
that the Government should focus on the importance of ensuring
everyone will have a good pension. This means devising a system
that not only extends affordable private pensions to those who
do not have them but also protects existing workplace provision
and the wider pensions market. It should also include incentives
to encourage wider participation, higher contributions, and lower
costs over the medium term.
4. PENSIONS REGULATION
AND THE
STAKEHOLDER EXPERIENCE
4.1 Our proposals for Partnership Pensions
are designed to address the problems of the current regulatory
framework for pensions, and the lessons of Stakeholder pensions.
The Stakeholder Pensions Experience
4.2 Stakeholder pensions have not led to
the hoped-for increase in saving. More than 80% of Stakeholder
schemes set up by employers have no participating employees. There
are three main reasons:
— the absence of mandatory employer
contributions mean low incentives to save;
— pension providers and financial advisers
are often unable to recommend Stakeholder pensions to low and
moderate earners because the means-tested Pensions Credit makes
it hard to satisfy the "suitability" requirements of
the FSA's current regulatory regime;
— the charge cap is set at a level
making it difficult to reach out to the target group of savers.
In addition, high regulatory capital requirements,
especially with regard to sterling reserves for future expenses,
also make Stakeholder pensions less economic. The fear, possibly
misplaced, of retrospective Financial Ombudsman Service intervention
has also made providers and advisers cautious in promoting Stakeholder
pensions.
4.3 The price cap was itself set in a way
which did not follow the customary practice—applied by law
in some utility markets—of evidence gathering, extensive
economic analysis and open consultation conducted independently
of Government, subject to a clear timetable and with a right of
appeal to the Competition Appeal Tribunal.
4.4 Further, although the 2002 Sandler Review
advocated a significant relaxation of sales regulation for Stakeholder
products, the FSA's stakeholder regulations—Basic Advice—have
allowed very few pension sales, as the FSA's own research has
shown.
4.5 Finally, the slow pace of decision-making
and the lack of clear responsibility among the public agencies
involved diminished the likelihood of both a successful public
policy solution and the ability of companies to react commercially.
The Stakeholder brand was damaged as a result.
5. LESSONS FROM
STAKEHOLDER PENSIONS
FOR THE
DESIGN AND
REGULATION OF
A NEW
PRIVATE SAVING
SCHEME
5.1 At present, responsibility for the regulation
of pensions is fragmented and involves several different Government
departments and regulators, including the Treasury, the Department
for Work and Pensions, the Financial Services Authority, the Pensions
Regulator, the Financial Ombudsman Service and the Pension Protection
Fund. This makes for poor policy co-ordination, operational difficulties
(particularly for employers), and does little to encourage pension
saving.
5.2 The Stakeholder pensions experience
provides many lessons for both the design and regulation of any
new private saving scheme. We have incorporated these into our
proposals for Partnership Pensions.
Lessons for the Treasury on the Design of a New
Private Saving Scheme
5.3 The policy and regulatory framework
should be designed as an integrated whole. This will allow simpler
and less costly regulation. In particular, reduced reliance on
means-testing and the use of an employer contribution will together
mean that the current "suitability" requirements are
unnecessary. The regulator should be set clear, and fully joined-up,
objectives that serve the consumer interest by encouraging saving
as well as ensuring consumer protection.
5.4 The Stakeholder experience shows the
dangers of a charge cap. If one is considered necessary, it should
be adopted only after an adequate period of market monitoring
and on the basis of a sound, independent, analysis of the nature
of any market failure and the costs and charges faced. This should
be the responsibility of a regulator with a remit closer to those
of the economic utility regulators.
5.5 The wider regulatory environment is
also relevant. It is important that neither the regulatory regime
for credit and borrowing, nor that for pension saving outside
the new scheme, should discourage saving or undermine existing
workplace provision and the wider pensions market.
5.6 Finally, getting the regulation right
at the start is essential. The Government's White Paper should
set out its expectations for regulation and, if necessary, foreshadow
any necessary changes to the Financial Services and Markets Act
2000, as well as pensions legislation.
Lessons for the FSA on the Implementation of a
New Private Saving Scheme
5.7 The Pensions Commission's assumption
that pension saving is the right thing to do for most people is
to be welcomed, and should be the starting point for the FSA's
approach to any new system. This will avoid the problems of Stakeholder
pensions, where the emphasis was on guarding against the risk
of "unsuitable" sales.
5.8 Further thought should be given to the
rate at which prudential requirements are set. The existing sterling
reserve requirements for future expenses are not risk-based and
add unnecessarily to the cost of pensions.
5.9 The regulator should work with the Government,
trade associations and consumer groups to ensure that consumers
are provided with the information and decision-making tools (eg
decision-trees) they need in order to decide on whether or not
to opt out of the new system, whether to save at the minimum level
or higher, and whether to accept a default fund or choose one
of the alternatives.
5.10 Very careful consideration should be
given to the impact of the new system on existing workplace pensions
and the wider pensions and savings market. Many on middle to high
incomes will need to make quite substantial additional pension
provision if they are to avoid a large drop in their expected
living standards in retirement. This argues strongly against introducing
any successor to the current "RU64" rule, which the
FSA rightly has recommended abolishing.
5.11 The regulatory regime for the new system
should take full account of the Government's Five Principles of
Good Regulation. It should be proportionate, accountable, consistent,
transparent, and targeted.
6. CONCLUSION
6.1 The Pensions Commission's proposals
and the forthcoming White Paper offer a tremendous opportunity
to close the Savings Gap and achieve financial security in retirement
for all. We urge the Government to seek a comprehensive solution;
one that will require bold reforms of means-tested pension benefits
and a new approach to regulation, as well as the creation of a
new model for private pension saving.
6.2 Whatever decision is ultimately made
about the future of private pension saving, action taken now to
address the weaknesses within the current Stakeholder pension
framework would be a useful first step towards helping more people
save for retirement.
March 2006
Annex A
THE FUTURE OF PENSIONS REGULATION
Pensions regulation should move with market
developments, ensure that all stakeholders' interests (especially
those of employers) are properly represented and that the overall
effectiveness of the policy framework is evaluated and modified
in the light of experience. The current framework does not achieve
this.
The ABI therefore proposes the establishment
of a new statutory body, the Retirement Income Commission. This
would act as an economic regulator, as well as carrying out a
range of monitoring, supervisory and educational functions. The
RIC's role, aspects of which would be relevant in relation to
an NPSS model as well as to Partnership Pensions, would include:
— Economic Functions: The Commission
would apply the disciplines of economic regulatory analysis to
the pensions market. In the first instance, it would focus on
monitoring costs and charges with a view to assessing whether
the market is functioning properly. But the RIC should have a
power to recommend the setting of a price cap as a safeguard measure.
Any such cap would be set in line with best practice—a full
understanding of the costs faced and the nature of the market
failure, full and transparent consultation, and opportunities
for appeal to the Competition Appeals Tribunal.
— Monitoring Functions: The Pensions
Commission's Second Report sets out a compelling case for the
establishment of a small monitoring body to review the effectiveness
of the pensions settlement as a whole and to provide authoritative
independent advice to Government on necessary reforms. We believe
that such a strategic role should be an essential component of
the RIC so that it can analyse the interplay between state and
private sector provision, providing early warning of a failure
to meet participation or contribution targets. The Commission
could also provide authoritative research, provide thought-leadership
and evaluate best practice. A further key monitoring function
would be the tracking and analysis of employer behaviour—ensuring
compliance with any compulsory employer matching contribution,
and checking that the new pension system did not lead to a "levelling
down" of existing workplace pensions.
— Operational and Consumer Protection
Issues: The RIC should undertake the small number of central operational
functions necessary to sustain Partnership Pensions, such as running
the "carousel" of providers to be used where employers
do not make an active choice of the default pension provider.
— Promoting Saving: The RIC would provide
a disinterested authoritative public voice promoting the need
for individuals to save for retirement. It would draw on the experience
of the New Zealand Retirement Income Commission, a small public
body which, through innovative web and other media resources,
has helped improve financial literacy and retirement saving.
Simplifying the Landscape
It may also make sense for the Commission to
subsume the supervisory and solvency functions of the Pensions
Regulator as a means of simplifying the overall regulatory environment.
On the same principle, the RIC might also take on the operational
mantle of the Pensions Protection Fund. This would have a number
of advantages:
— More rounded decision-making with
a single body.
— Better assessment of the impact of
pensions policy as a whole on employers.
— Economies of scale in operational
functions.
Reform also provides an opportunity to review
and clarify the boundaries between the Financial Ombudsman Service
and the Pensions Ombudsman.
We do not, however, propose that the RIC should
take on the prudential role of the FSA in relation to financial
services companies. Solvency regulation should happen once for
companies, rather than being fragmented between different regulators
on a product basis.
| 
