Memorandum submitted by Genetics Group,
Faculty and Institute of Actuaries
1. This memorandum has been prepared by
the Genetics Group of the Social Policy Board of the Institute
of Actuaries and the Faculty of Actuaries, the professional bodies
for actuaries in the United Kingdom. In view of the short time
available for its preparation, this memorandum has not been approved
by the Social Policy Board, or by the Faculty and Institute Management
2. UK insurers offer the best terms they
can and the central plank of consumer protection is, in fact,
the informal one of competition. Insurers will not recruit or
retain customers if they are more expensive than their rivals.
This does not result in every insurer making an attractive offer
to every member of the population. But it means that most members
of the population can meet their insurance needseven if
it is in some cases by recourse to specialist niche insurers and
on specially calculated tariffs.
3. There are however segments of the population
that effectively cannot contract insurance because their risk
of ill-health or death is so high as to render the contract impractical
or its price unaffordable.
4. An insurance market, in the absence of
price controls, also has the shortcoming (arguably) that disadvantaged
groups are not subsidised.For example, known cardiac risks
will generally pay more for life insurance (although their health,
employment status and financial plight may mean they are less
able to pay).
5. In the absence of regulation insurers
could be expected to be interested in genetic testing. Firstly,
if it were possible to offer better terms to customers who had
a certain result in a given genetic testthen this would
advantage such customers and the operation of a free market would
facilitate such a development. We do not have to explore the merits
and demerits of this however, since UK insurers have committed
to not requiring customers to take tests. Also, there is no test
currently available that would make such a stark differentiation.
6. The more significant issue is that insurers
instinctively (if they are operating on the competitive market
model) would ask customers whether they have undergone genetic
tests and what the results were. This is a natural extension of
the conventional underwriting process whereby customers effectively
declare all information that they (the customer) possess which
could affect the insurers' assessment of the risk.
7. Clearly, if a genetic test were to become
accessible and affordable and it allowed 30 year olds to identify
that they would die before age 40then it would be impossible
for insurers to operate without taking such a test into account.
If they did not, it would be tantamount to offering an 80 year
old life assurance for a 30 year old's price.
8. It is understandable therefore that insurers
are reluctant to surrender all rights to use the results of genetic
testing. The Committee could perhaps explore whether insurers
are concerned to protect such rights to contend with future genetic
science, as yet unknown, or which of their concerns are real live
9. In 1997, the Human Genetics Advisory
Commission, in a report entitled "The Implications of Genetic
Testing for Insurance", recommended that "a requirement
to disclose results of specific genetic tests would only be acceptable
where a quantifiable association between a given pattern of test
results and events actuarially relevant for a specific insurance
product has been established.
10. The concept of "actuarial relevance"
has been carried forward into the criteria of the Genetics and
Insurance Committee (GAIC). Submissions from the insurance industry
to the GAIC to request permission to make use of particular genetic
tests, must demonstrate not only that the test in question has
clinical relevance, but also that it is actuarially relevant.
11. In one sense, all information pertaining
to the risk propensity of policyholders and prospective policyholders
is relevant from an actuarial point of view. The actuary should
take into account all available information about the group of
people being insured, in order to recommend a fair premium rate
for the group, and, subsequently, to ensure that the insurance
company sets up appropriate provisions for the cost of the claims
which can be expected to arise. This is necessary, both for the
protection of the insurance company (and hence other policyholders)
and for there to be fair treatment between different groups of
12. The particular characteristics of individuals
proposing for insurance can, to the extent that they are required
to be disclosed, be taken into account by the insurance underwriter,
in deciding on the premium rate to charge and on whether to accept
the proposal on standard terms, to impose a premium loading (because
of a perceived higher level of risk), to allow a discount (because
of a perceived lower level of risk) or to decline to insure.
13. Age and sex will almost always be considered
as actuarially significant characteristics, but so, according
to the type of insurance, may be occupation, place or residence,
current health condition, lifestyle, smoking habit, etc. Generally,
a characteristic will only be used for setting differential premiums,
or for informing the underwriting decision, if the presence or
otherwise of the characteristic is deemed, on the basis of the
evidence available to the insurer, to make a significant difference
to the risk propensity, ie to the probability of a claim or the
likely incidence of a claim.
14. Simple questions are usually asked in
a proposal form about the proposer's state of health, past illnesses
and treatment and, in some cases, about the age at death and cause
of death of close relatives. In cases where this preliminary information
suggests indicators that there could be a higher than average
risk propensity, the underwriter may ask for a report from a doctor,
or request a medical to be carried out. This would be more likely
to occur for larger than average sums assured, which clearly expose
the insurance company to a risk of a greater level of loss.
15. In practice, even most of the cases
where some risk factors are present will be insured on standard
rates. It is believed that some 95 per cent of proposals for life
insurance are accepted on standard terms and only less than 1
per cent of proposals are declined.
16. The result of a genetic test would be
significant from an actuarial point of view ("actuarially
relevant" in GAIC terminology) if the knowledge of a particular
result from the test would make a noticeable difference to the
actuary's assessment of the risk propensity, of the order of magnitude
which would be expected to lead to a premium loading or a premium
reduction compared to the position without knowledge of the test
result. In a more extreme case the test result might lead to acceptance
of a proposal which might otherwise be declined or rejection of
one which might otherwise have been accepted.
17. There is no unique criterion (or set
of criteria) for saying whether a test result is significant.
Genetic test results (and other disclosures which might be required
of the prospective policyholder) may come anywhere in a spectrum
from those which provide no information at all to affect the underwriting
decision to those which introduce a totally new perspective on
the risk from the insurer's point of view. As with all underwriting
considerations, the issue will be of greater importance to the
insurer if the sum assured is well above the average.
18. The GAIC have determined a trigger for
actuarial relevance based on extra mortality risk of 50 per cent
(ie the chance of dying in the relevant period is increased by
more than a half again of the normally expected chance). This
probably corresponds broadly with the level of extra mortality
risk which would normally lead most insurers to consider imposing
an extra premium.
19. In order for this level of extra mortality
to be demonstrated, the results of appropriate research projects
on subjects with different test results must be sufficiently robust
to demonstrate that the presence of particular test results in
a group of individuals can be expected to push up their expected
probability of death by 50 per cent or more. This is in comparison
with an average group of individuals, for whom there is no particular
reason to suppose a high risk propensity.
20. In practice, the availability of such
a test result to the insurer may not lead to a higher premium
being charged than otherwise. This could be 1) because other information
on the proposal form, or from a medical report, would have led
to that level of loading anyway 2) because the sum assured is
small, or 3) because the company's approach to the mutualisation
of risk is such that it is prepared to accept some proposers with
significant additional mortality risk on standard rates, as part
of its broad underwriting strategy (most likely subject to some
upper limits on sums assured).
21. The extent to which insurance companies
mutualise risk is affected by what they perceive society's attitude
to risk-sharing to be. Social security schemes usually involve
a very high level of mutualisation of risk across the whole community.
In the interests of keeping premium rates down for the majority
of policyholders, and in order to remain competitive relative
to others, insurers usually tend to restrict the mutualisation
of risk to within groups of policyholders with broadly similar
characteristics, eg the same age, sex and without any unduly high
additional risk characteristics.
22. Under the GAIC process insurers must
demonstrate that any genetic test they use is "actuarially
relevant" as discussed above. This process has produced some
benefits for customers and the insurers:
(1) It prevents insurers misinterpreting
less understood genetic testsan insurer might have, for
example, declined cover because it had "doubts".
(2) It has brought clinicians, geneticists
and insurance professionals together and reduced somewhat the
fears the medical profession have (for example that patients become
discouraged from undergoing tests).
(3) It has allowed time for realisation of
the as yet limited implications of genetic testing in insurancethe
impact has not been dramatic or revolutionary in nature.
23. The GAIC process should not however
be regarded as solving all the issues:
(1) It does not lead to prescription of how
an insurer will price insurance for a customer who is affected
by a GAIC-approved genetic test. Indeed it would seem illogical
if it didsince in non-generic matters insurers do not generally
have their actions constrained in such fashion.
(2) It does not mean the following groups
of customers are not "disadvantaged" in insurance:
People with genetic conditions
with a GAIC approved test;
People with genetic conditions
revealed to insurers in other ways (eg family history) although
there is no GAIC approved test;
People with a medical condition
that is not considered genetic.
(3) In effect the GAIC process seems to protect
a small minority of customersthose whose genetic condition
is known to them via a test but which has no other revealed symptoms
of which an insurer would be aware.
24. It is evident that modus operandi
of insurers can cause considerable distress to people with deteriorated
health. At the very time when they require to replace income for
themselves, or provide for loved onesthe commercial actions
of insurers can seem unhelpful, awkward and even discriminating.
But this is by the nature of the UK's competitive market model.
25. Insurers have not had explicit social
obligations placed upon them and they cannot be expected to follow
some or other norm unless some framework is provided by regulation
(which an insurer will know its rival competitors are also subscribing
26. Such social obligations of course could
be placed upon insurers and there are different examples in different
insurance markets around the world. But it is generally not a
simple task. The more onerous the obligation, the more rigorous
needs to be the framework and enforcement: especially to ensure
such insurer meets its social obligation in equal measure to others.
27. The moratorium on insurers using certain
genetic tests results and the ABI's voluntary abstention for genetic
tests for normal-sized mortgage-related life insurance are examples
of "lightweight" social obligations. As observed earlier
they do protect a small minority of the population but the social
protection afforded is not targeted in any coherent way.
28. It is clear that patient interest groups
will campaign very cogently and convincingly on behalf of their
members and the prohibition of insurers using certain genetic
tests as an act per se may be of small benefit. But it will not
solve the more deep-seated problem of economic deprivation associated
with deterioration in health. While working towards one social
policy objective, solutions proposed must not render inoperable
the competitive market model for insurance.
29. We would close our memorandum with a
useful example and pose a question. It is emerging that genetic
tests can indicate a higher propensity for Alzheimer's Disease
in some individuals. How then should society proceed if it wants
to foster the private long-term care insurance market? If society
permits genetic evidence to be used by insurers it could lead
to a disadvantaged group (at least by price comparisons). On the
other hand a prohibition could mean only the "at risk"
take out insurance, which becomes expensive and then a wider market
does not prosper. There are no easy answers we would suggest.
30. The actuarial profession in the UK expresses
opinions on public interest topics, even though these may conflict
with the views of some individual members of the profession or
with the views of commercial clients in the industries which actuaries
serve. A statement of the profession's position on Genetics and
Insurance is posted on the profession's website (www.actuaries.org.uk).
A copy of this Position Statement is attached as an Annex to this
31. The actuarial profession would of course
be very honoured to give further help to the committee as it tackles
this difficult matter.
22 January 2001