APPENDIX 22
Memorandum submitted by NatWest (CP 32)
NATWEST PAPER ON THE INSURABILITY OF WELFARE
BENEFITS
This paper examines briefly the factors which
affect the insurability in the private sector of benefits which
are currently provided by the state.
SUMMARY OF
MAJOR WELFARE
BENEFITS
— Widow's/widowers' benefits
— Benefits relating to disablement,
care, attendance
— Housing related benefits
The list covers a wide range of benefits, payable
in different circumstances and for varying periods of time. For
example, qualification conditions for a benefit may fall into
one of the following categories or combine elements of each:
— automatic payment in specified circumstances
eg, child benefit;
— payment depends on contributions
or employment history eg, pensions;
— benefit is means tested eg, income
support.
There may also be age limits for claimants and/or
limits on the length of time a benefit will be paid.
Within these categorisations, benefits can be
divided into those which are designed to help individuals maintain
a social acceptable minimum income over a period of time and those
which are designed to provide money when there are significant
extra needs as a result of a particular event.
Each benefit should be examined separately but
there are a series of common tests which can be applied to establish
the potential for effective private insurance.
PROFITABILITY
At the most basic level, to make an underwriting
profit, the premiums charged by an insurer must exceed the total
of claims, expenses and distribution costs. In some cases, it
is possible to write insurance business at an underwriting loss.
In other words, claims paid plus expenses exceed premiums charged.
This is because a profit can be made from investing premiums eg,
if there is a long average period before claims payments are made.
The existence of active insurance markets suggests that in many
cases, people happily accept the certainty of the small loss (the
premium) in exchange for the prevention of a large loss which
may or may not arise.
Providing value also depends on a straightforward
but safe regulatory environment in which sales and administration
costs can be kept to a minimum. Tax incentives can also reduce
the effective premium to make insurance more attractive.
The financial objectives of state provided insurance
will be different from the private sector. Benefits are paid for
by the state using money raised from taxation, including National
Insurance. There need not be a direct relationship between the
National Insurance "premium" and the level of benefit
as there would be with private insurance. The state is therefore
able to introduce greater cross subsidy between individuals and
types of benefit. The financial objective of the state may be
to provide an adequate and politically acceptable welfare system
without imposing unacceptable costs on the Exchequer.
INDIVIDUAL CHOICE
From the point of view of the individual, the
basic profit requirement above means it is statistically unlikely
that they will profit financially from the insurance. They will,
however, be prepared to buy it if they value the security it provides
more than the cost of the premium. This depends firstly on awareness
of risk and secondly on attitude to risk. These factors vary between
individuals and can be influenced by education and by marketing.
Attitude to moral hazard may also play a part and is discussed
later in this paper.
Whether people prefer to insure privately rather
than through the state will also depend on confidence in the provider.
Many people believe that state benefits are more securely guaranteed
but regular alterations of benefit conditions and calculations
demonstrate this is not always the case. Benefits from private
insurance depend on legally binding contracts and, in some cases,
investment performance. It is also possible to construct compensation
schemes to protect benefits if the insurer becomes insolvent.
Given the choice different people will want
to insure different benefits and at different levels. This is
not a choice which exists within state welfare provision and it
may be difficult to achieve in a privatised system. Assuming there
is a socially acceptable minimum standard of living, there would
still have to be a safety net for people who had chosen not to
insure. If provision of the safety net is left to the state, it
will effectively remain the welfare provider for those who choose
not to insure or who cannot get insurance. The incentive would
be not to insure, knowing that the state would still be there.
Compulsion may therefore be required if insurance
replaces welfare and it must be at or above the level required
to provide the socially acceptable minimum. The compulsion does
not have to force people to insure privately. People could be
allowed a choice between state and private sector as the provider,
as currently exists for contracting out of SERPS. A third choice
might be to allow self insurance as an alternative ie, putting
aside a certain amount of savings which cannot be used for other
purposes.
CONTROL OVER
THE INSURED
EVENT
Eligibility for state welfare is unlikely to
depend on the reason that the individual needs help. Judgements
are not based on whether "it's your own fault". In contrast,
a general principle which applies to private insurance is that
the insured must not bring about the insured event.
There are some private insurances eg, redundancy
cover and maternity cover where this principle would be hard to
judge. If it is impossible or impractical to judge whether the
insured has brought about the event, can the benefit be insured?
In order to achieve the intended peace of mind, the individual
needs some certainty of what is covered.
If an event occurs which is not a valid claim
but still results in a welfare need, how will the need be met?
There could be an uninsured loss pool as exists for motor insurance
or an industry reinsurer like Poole Re. If insurance is compulsory
for all, this would provide a large enough pool to give more predictable
incidence of claims and may allow wider cover. As with state welfare,
good risks would be paying more than the strict cost of insurance
and poor risks would get cheaper cover than would otherwise be
possible.
A risk which cannot be offset is the risk of
economic recession or a very strong currency which may, for example,
cause unemployment to rise across the country—a possibility
a government but not an insurer has some control over.
FUTURE UNCERTAINTY
The future uncertainty which affects the individual,
also affects insurers and the state. Of the three, the state has
the most control because it effectively defines welfare. The changing
expectations of society shape the development of the various benefits.
Insurance may be all about risk but insurers
don't like to accept unquantified risks. It must be possible to
quantify risk so that future claims incidence and cost is predicted
with reasonable accuracy. This requires stable external influences
during the period of insurance. It should be possible to base
future insurance claims on past welfare claims but it is not straightforward—welfare
conditions are frequently altered and costs are continually rising.
If the definition of incapacity can be changed again in future,
the risks become unquantifiable. The period of insurance will
be limited to the number of years ahead for which claims can be
comfortably predicted. The public, however, would prefer premiums
to be fixed for a long period. They may be particularly reluctant
to accept uncertain future costs if they are being set by insurers
rather than elected representatives.
POOLING RISK
It is necessary to build an acceptable portfolio
containing balance of good and bad risks so that the correct premium
can be calculated. Taking a benefit such as unemployment payments,
different ages, occupations and geographical areas carry different
risk. It may be possible to assess each individual using all available
criteria and calculate the risk they bring to the pool. However,
this would be an expensive process in itself and it would result
in excessive and unaffordable premiums for some groups, who would
find themselves uninsurable in a free market.
The state could assume the risk for all those
who were unable to buy insurance at a level which they can afford.
To encourage take up of insurance, National Insurance rebates
could be available for individuals who choose to take out private
insurance. Such individuals would be unlikely to effect private
insurance if the costs were much above the rebate level. The people
remaining with the state would therefore be the ones with the
highest premium and highest risk. In other words, as a provider
of insurance the state would be selected against.
Another option is to pool all the risks, charging
everybody the same premium. For this to work, there could be no
choice of insurer and the insurer could not turn down any individual.
Otherwise, for example, insurers providing unemployment cover
would actively market their products only to particular age groups,
industries and geographical areas with low unemployment.
A no claims bonus system may provide a compromise
between individual risk assessment and national level pooling
and could also reduce claims. Rather like with motor insurance,
the incentive would be not to claim unless it is unavoidable.
Another option would be to calculate the range
of risk which is "normal" and for only that level to
be insured. The state would then top up either the premiums or
the benefits when the risk exceeds the normal level.
MORAL HAZARD
AND FRAUDULENT
CLAIMS
There is some concern that the people who take
out insurance may be those who most think they will need it. Similarly,
the behaviour of someone who is well insured may be different
from an uninsured person. They could be more willing to accept,
or even seek out the risk they have insured, knowing that they
are protected from its consequences.
This seems unlikely in welfare insurance where
benefits will be low and there is little to be gained by claiming.
Of course, moral hazard is also present when the state is the
provider. It is well known that there are fraudulent claims within
the state welfare system and this is something the private sector
would need to minimise, at least to the point where the costs
of doing so don't outweigh the benefits. It is estimated that
fraudulent claims amount to £4 billion a year which is approximately
double the expected total cost of the pension transfer and opt-out
review each year. Attitudes to fraudulent claiming have a part
to play and they may vary between the public and private sectors.
Research in this area would be useful to establish whether fraud
could be reduced in the private sector.
An element of savings which would be drawn on
before a benefit began to be paid could act as a deterrent to
claiming. Giving people individual accounts would also focus attention
on the cost of welfare and it is possible to design a system which
incentivises people not to claim because they will be paying the
first part of the claim themselves.
ADMINISTRATION
It may be possible to separate the administration
from the insurance. In some parts of the World eg the USA, private
administration is used for some state paid welfare. If the main
benefit of privatisation is seen as administrative efficiency,
this may be achievable without insurers taking on the risk. In
terms of premium collection, the PAYE and NI systems operated
largely by employers could be considered. Certainly, if contributions
are compulsory then one way to ensure they are paid is deduction
at source. New clearing house systems may be a viable alternative
to divert the correct amounts to each insurer so that employers
do not have to forward premiums to a number of providers.
Claims administration costs can be reduced if
deferred periods are used for benefits such as unemployment or
incapacity where the claimant may recover.
FIT WITH
OTHER SAVINGS
It may be possible for savings to cover some
of the risks. For example, the first two months of unemployment
or incapacity. This would work like an excess in motor or household
insurance to reduce the numbers of small claims. For this to work,
the insured would have to have restricted access to the savings
so that they can only be used in specific circumstances. When
the funds are used up, the claim would start paying. This would
also allow the customer virtually instant access to money while
giving the insurer a couple of months to check the claim. A given
level of savings could only cover one insured event unless the
claims are mutually exclusive.
When the claim is over, the insurer could continue
to pay until the savings are restored or premiums could rise to
cover the full insured amount until the savings had been restore.
Alternatively, could the savings account be overdrawn and if so,
would it be the state or private sector who underwrite any bad
debts?
TRANSITION
It is important that any new system does not
exclude people but there are some existing welfare claimants no
insurer would be willing to take on unless they were subsidised
by the state.
There have already been trends towards the private
sector for medical, incapacity and old age care but a sudden move
to the private sector would be socially and politically difficult.
Partnership schemes will encourage people to
move.
SUMMARY
There are clearly some difficulties which arise
if welfare insurance is provided by the private sector. For some
forms of cover, it should be possible to overcome these problems.
The market cannot provide a fully inclusive
system without some state involvement. It may be necessary for
the state to retain two roles. Firstly it can assume elements
of the risk which are not insurable because the incidence or cost
cannot be reasonably predicted. An example of this might be long
term unemployment which is dependent on economic factors which
insurers cannot influence (although short term unemployment might
be insurable).
Secondly, the state could help with costs of
insurance for groups such as the elderly and the sick. For these
groups higher premiums (essential to recognise the higher risks)
will often be unaffordable. Unless the state remains the provider,
such people would need assistance in order to get private insurance.
This could take the form of support for the individual to enable
them to pay the premiums or payments direct to insurers to meet
premiums which exceed a certain level. This effectively separates
out the redistributive welfare element of National Insurance from
the insurance element. The redistributive element may benefit
from simplification if the insurance element is privatised. Of
course, over a period of time, the existence of adequate insurance
against illness, redundancy etc will reduce the numbers of people
whose income is reduced to the point where they cannot afford
insurance.
Alternatively, if homogenous products were developed
and insurance was compulsory, a national average premium may be
possible. Insurers would be compelled to accept people with no
selection criteria. This basically transfers the redistributive
functions to the private sector. People who bring a low risk to
the pool would be paying too much and people at high risk would
be undercharged.
Some element of compulsion will be necessary
to ensure people take out at least the minimum adequate cover.
Within this, they may be able to choose between the state and
the private sector or elect to self insure by setting aside savings
which cannot be used for other purposes.
Grouping similar insurances together may allow
more efficient provision. For example, income protections insurance
could cover illness, disability, accident, and redundancy. Such
cover may also fit well within pension provision. Grouping risks
will save administration and underwriting and provide a more diverse
portfolio of risks.
There needs to be a seamless fit between the
state and private sector and a clear division of responsibility.
This requires a degree of standardisation of insurance products,
at least for the minimum adequate level of cover. To allow risks
to be assessed with a reasonable amount of certainty, policy conditions
will have to be immune from changes to welfare levels, eligibility
or structure during the policy term. If there are such changes
during the period of insurance they would have to be covered by
the state.
Segments which have the fewest obstacles are
those where the level of risk varies the least between individuals,
is reasonably predictable and is largely outside the control of
the individual. These might include risks associated with:
Less appropriate are risks associated with:
— Maternity, child benefit, one parent
benefit
— Some types of unemployment
— Income Support, family credit
Further research is required into some of the
options discussed in this paper so that the most appropriate approach
can be decided.
15 June 1999
| 
