Get documents about "Technology Innovation Trends Insurance Claims Adjusting
W
Description
Technology Innovation Trends Insurance Claims Adjusting document sample
Document Sample
EU Project INTEREST HPV1-CT-2001-60024
INTER EST
INsurance, TEchnological Risk and Emerging
Science and Technology policies
Report of the first workshop held at the PSI Conference Centre,
London
28 February – 1 March 2002
Report prepared jointly by:
Kristina Dahlström, Policy Studies Institute
Jim Skea, Policy Studies Institute
April 2002
Policy Studies Institute
100 Park Village East
London NW1 3SR
www.psi.org.uk
CONTENTS
Page
Summary………………………………………………………………….…..3
1. Introduction
- aims of project………………………………………………….….8
- aims of workshop……………………………………………….…8
2. Unde rlying Concepts
2.1 Insurance, Risk Management and Culture, Walter Stahel………………..10
Discussion………………………………………………………………...13
2.2 When is a Risk Insurable?, Gerry Dickinson……………………………..15
Discussion………………………………………………………………...19
2.3 Innovation and Insurance, Bas de Laat…………………………………...21
Discussion………………………………………………………………...32
2.4 Sustainable Development, Risk, and R&D Policy, Jim Skea/Uno Svedin..33
Discussion……………………………………………………………...…40
3. Case Studies
3.1 Insurance and risk management tools to enhance security and reliability for
innovation: Nymphea Water, Jean Michel Ribouleau…………………….42
Discussion…………………………………………………………………51
3.2 Regulating new technology: The Case of GM Crops, Simon Joss with Malcolm
Eames……………………………………………………………………...54
Discussion…………………………………………………………………64
4. Conclusions
- final discussion…………………………………………………….70
- key points arising from the workshop……………………………..70
Annex 1: List of Participants……………………………………………….73
Annex 2: Workshop Agenda………………………………………………..74
2
SUMMARY
The INTEREST Project
1. The INTEREST project addresses the role which the insurance sector and
insurance-based mechanisms can play in relation to innovation which promotes, or
challenges, sustainable development. The insurance sector and insurance-based
mechanisms may have a greater role to play in managing the risks associated with
innovation, either by shifting the boundaries between private sector and public sector
management of risk, or by developing new innovative arrangements in which
insurance and regulatory mechanisms play complementary roles.
2. The objectives of the first INTEREST workshop were to: a) identify the past and
present role of insurance in relation to innovation which promotes or challenges
sustainable development; and b) to identify opportunities for promoting sustainable
development through innovation by adjusting or sharing the responsibilities of private
sector actors, regulators and policymakers.
3. Four conceptual papers, outlining key concepts and ideas, underpinned the
workshop, and two case studies, on the NYMPHEA Water Project and GM crops
grounded the discussions.
Insurance and the Insurability of Risk
4. Opportunities as well as risks must be considered. Risks must be placed in their
cultural context. In the aftermath of September 11, it is apparent that the greatest risks
come not from nature but from society itself, and more emphasis must be placed on
imagining maximum possible losses.
5. Insurance can influence technological choice. Insurance – or the lack of it –
provides clear market signals as to the level of risk inherent in different technologies,
as future costs are considered in calculating premiums. Insurance not only
compensates for loss but can also, in some instances, be designed to prevent losses
from occurring in the first place. Insurance could play a role in selecting sustainable
technologies, and can be regarded as a potential policy tool.
6. The capacity of an insurer to provide insurance depends on regulatory and legal
limitations as well as pricing issues. The events triggering insured losses must be
sufficiently well-defined. An insurer must be able to estimate with a sufficient degree
of accuracy the likelihood and severity of losses. Those most likely to suffer losses
should not disproportionately seek insurance (the problem of adverse selection). The
existence of insurance protection should not cause the behaviour of insured parties to
change the frequency and severity of insured loss in an unpredictable way (moral
hazard).
3
7. Insurers must have the capacity to provide sufficient risk transfer. This, in turn,
depends on the financial resources of the insurance market in relation to the scale of
potential losses. The scale of potential losses can be reduced by preventative action
and risk management support , activities which can also increase the set of insurable
risks.
8. There are constraints on the demand for insurance. If prices are set too high,
insurance may become unaffordable, affecting certain individuals disproportionately.
Governments could play an increasingly important role in insurance, by providing
compensating subsidies or for example preventing or reducing the risk of war and
terrorism.
Innovation and Insurance
9. Innovation processes are non-linear and are characterised by the interactive
involvement of a multitude of actors. Innovations tend to be path-dependent, leading
to the establishment of often irreversible technological trajectories. The more
established a technology becomes, the more difficult it becomes to influence it. From
the insurance perspective, it is important to identify those technologies where early
closure of socio-technical options occurs.
10. There is no data on innovative technologies to enable probabilistic assessme nt and
their impacts may be highly uncertain. Traditional actuarial approaches cannot be
applied. In this situation there may be three opportunities the insurance sector to
engage with the management of innovation by: a) engaging in foresight/horizon
scanning activities; b) engaging earlier in the R&D process to understand better future
risks; and c) providing experience and insight to stimulate public policy debates.
Sustainable Development, RTD Policy and Insurance
11. To make further progress in bringing sustainable development principles into
policy- making, there is an apparent need to connect the local, regional and global
scales and to articulate better the interface between policy and science.
12. The insurance sector can contribute to sustainable development policy and tools
by: a) participating in fora that help identify sustainable development policy needs
and RTD priorities; and b) disseminating knowledge of societal and environmental
risks and insurance-based mechanisms for dealing with them. In this way, the
insurance sector could gain insights into precautionary approaches leading to better
loss prevention, improve its reputation and help to expand the set of insurable risks.
The NYMPHEA Case Study: Technological Performance Achievement Guarantees
13. The first case study assessed the scope for a new product – the Technological
Performance Achievement Guarantee. It drew lessons from the Nymphea Water
project, which attempts to tap water resources from freshwater resurgence on the
4
seabed. The project relates to sustainable development in different ways, in that it
could have a positive impact on local economies and society while mitigating the
adverse environmental impacts associated with water scarcity.
14. The possible underperformance of a new technology entails financial
consequences that could affect the business plan, endanger third party involvement
and undermine further development. The source of underperformance is an important
factor determining the types of insurance available. The lack of historical data
impedes the traditional actuarial approach to the detriment of the innovation process.
The actuarial approach needs to be replaced by a factual approach, based on industrial
and scientific risk assessment. New insurance products, underwriting beha viour, and
risk management tools need to be employed.
15. The use of the Technological Performance Achievement Guarantee would
increase the role of the insurer, who could become a pro-active partner in project
management rather than just a service provider. The insurer would then be an “active”
actor fostering innovative behaviour and businesses working towards success. The
insurer will not only cover the consequences of failure but will have the ability to
prevent project failure in the first place.
The GM Crop Case Study: the Respective Roles of the Private and Public Sector
16. Genetic modification is a contested innovative technology that potentially offers
not only significant benefits but also significant risks. Both would have important
sustainability implications. Genetic modification has been a contentious issue in the
scientific, political and public spheres ever since the first experiments with
recombinant DNA in bacteria took place in the early 1970s. Environmental, social
and financial risks need to be placed in the wider context of the risk-benefit debate.
17. The absence of a strict liability regime has arguably been an important factor
facilitating the introduction of GM crops into the European market. The protection
offered to developers by the GMO Deliberate Release Directive and the Novel Food
Regulation has essentially allocated much of the potential risk of any unexpected
hazards arising from these products to the public sector and society at large.
Nevertheless, novel liability problems relating to the GM contamination of non-GM
crops have recently begun to emerge with discussions on the Environmental Liability
Draft Directive.
18. There is no counterfactual to the current situation, i.e. a case study of how the
market would have operated with strict liability in place. However, there are
indications that the insurance industry would have been unwilling to underwrite the
currently unquantifiable risks associated with GM had strict liability had been
required. This case study raises important questions about the capacity of the
insurance sector to deal with broader, societal risks.
5
Key points arising from the workshop
19. There are two possible roles for insurance in relation to innovation and
technological development with sustainability implications:
Where there is a prima facie case that a new technology will contribute to
sustainable development, and there are barriers to its uptake in the marketplace,
insurance mechanisms can facilitate its adoption by sharing technological risks
and providing a more secure financial framework for developers. However,
insurance cannot address the entrepreneurial risks associated with innovation.
Where the sustainability or otherwise of new technology is contested, the role
explicitly or implicitly assigned to insurance by the State, through regulations and
the use of liability regimes, and the consequent implications for the allocation of
risks to different actors, will determine the degree to which insurance acts as a
mechanism regulating the rate and direction of technological change.
20. The role of insurance in any particular circumstance is dictated very much by the
legal framework in which the sector operates. The State may directly mandate or
prohibit insurance, or establish liability frameworks which indirectly determine the
role of insurance.
21. Insurance has a role to play in relation to financial risks and risks borne by
people, where a monetary value can be placed on those risks. (e.g. health risks).
Insurance cannot deal with pure environmental risks where ind ividuals or groups are
not directly affected and where there are no associated property rights.
22. Actuarial evidence is not strictly necessary for a risk to be insurable.
Quantification of risks is needed, but this could be based on an ex ante analysis of
risk exposure and the probabilities of insurable events occurring. This would require a
much closer relationship between insurers and insured parties. The insurer would
need to acquire and develop specific technological knowledge of the insured party‟s
activities. Expanding the set of insurable risk in this way would require an expansion
of insurance sector competences.
23. The boundary of state/private sector. Insurance mechanisms involve the
voluntary transfer of risk from one party to another. Both parties, the insurer and the
insured, will in principle feel themselves to be better off as a result of the risk
transfer. On the other hand, when the State transfers risk from one party to another
through legislation, or authorises certain activities in the absence of a liability regime,
some parties will involuntarily bear certain risks while others benefit. Where new
technologies expose some parties to risk, while offering wider and more distributed
benefits, the State may decide that the “positive externalities of innovation” outweigh
the potential costs to affected groups.
6
24. The likely role of insurance in relation to major “societal risks” (such as those
associated with GM crops) is limited and the State will always have a primary role in
regulating such risks. The insurance sector is unlikely to regard such risks as
insurable and would probably decline any opportunities to become involved.
25. The situation where the State offers no specific controls over a new technology
other than to define strict liability to third parties provides an interesting if
hypothetical counterfactual to the reality of regulation. Technological development
would then take place only in such a way that affected parties could be compensated
for any losses that they suffer and the risk were deemed to be “insurable”. This would
depend on the capacity of the sector to quantify the risks involved (whether by
actuarial means or ex-ante estimates of risk exposure) and whether the risks were
limited temporally and spatially. Some would see this hypothetical regime as
representing a true application of the “precautionary principle”. The way it would
take account of the distribution of risk across society might also be seen as
implementing the equity aspects of the sustainable development principle.
26. Using a strict liability regime as a hypothetical benchmark, highlights the fact that
the net effect of most regulation which limits liability, notwithstanding safeguards,
is to underpin innovation, sustainable or otherwise, by allowing developers to
appropriate the benefits of innovation while transferring risk to other parties.
27. The insurance industry shows little interest in “blue skies” research. The probable
role of insurance in the innovation process is “downstream”, particularly in relation
to the demonstration, adoption and diffusion of new technologies.
28. There is little practical connection between the insurance sector’s investment
activities and its underwriting activities. In principle, risk management via
underwriting could affect the “riskiness” and returns on investment (e.g. the
underwriting of flood risks could affect property investments) but in general the link
is likely to remain tenuous. The investment strategies of insurance companies tend to
be conservative and risk avoiding.
7
1. INTRODUCTION
Aims of Project
The INTEREST Project addresses the role which the insurance sector and insurance-
based mechanisms can play in relation to innovation which promotes, or challenges,
sustainable development. It forms part of the EU 5 th Framework STRATA Programme
(STRATegic Analysis of specific political issues).
New technological frontiers and sustainable development together will place new
demands on the insurance sector. The insurance sector and insurance-based mechanisms
may have a greater role to play in managing the risks associated with innovation, either
by shifting the boundaries between private sector and public sector management of risk,
or by developing new innovative arrangements in which insurance and regulatory
mechanisms play complementary roles.
The objectives of the project are:
1. to identify the past and present role of „insurance‟ in relation to innovation which
promotes, or challenges, sustainable development.
2. to identify opportunities for promoting sustainable development thro ugh innovation
by adjusting or sharing the responsibilities of private sector actors, regulators and
policymakers.
3. to establish how sustainable development innovation can be promoted using risk
management mechanisms from insurance, such as risk assessment, loss prevention
etc., in other domains.
4. to establish how and to what extent sustainable development innovation can be
promoted by the insurance sector using a wider set of tools, such as technology
assessment, to improve its capacity to manage novel risks.
5. to explore policy options for the complementary use of insurance-related and other
risk management mechanisms.
Aims of the workshop
This exploratory workshop addressed the first two project goals, bringing together a
mixture of participants from inside and outside the insurance sector. These included
science and technology policymakers, industry, sustainable development policymakers,
and members of the scientific community concerned with risk management and
technology assessment.
8
The workshop discussed the advantages and disadvantages of insurance/market-based
approaches to dealing with uncertainty, looked at differences and similarities across the
countries of the EU-15, and approaches used elsewhere. It addressed: the boundaries of
insurability and the nature of the risks that are, or might be, covered by insurance;
market-based approaches for dealing with uncertainty, such as international accounting
principles arising from the professional liability of accountants; existing insurance–based
risk management approaches including loss prevention; and new risk management
approaches, such as extending the limits of insurability beyond the non-probabilistic
realm by adopting the precautionary principle.
Discussions were grounded in two case studies. The first related to „performance bonds‟,
an insurance device intended to stimulate sustainable development innovation by
providing cover, to SMEs introducing new technologies, against technological risks
relating to under-performance, failure or damages to third parties. The second case study
related to the leading edge sustainability challenges associated with the adoption of GM
crops.
The case studies were underpinned by presentations outlining the key concepts and ideas
underpinning the INTEREST project. The presentations covered lessons from existing
insurance approaches to risk management, and what makes certain types of risk insurable.
Looking more widely, the management of risk in innovation was assessed, and the
broader links between sustainable development, risk and R&D policy were also
addressed.
9
2. UNDERLYING CONCEPTS
2.1 Insurance, Risk Management and Culture
Walter Stahel
Inte rnational Association for the Study of Insurance Economics (Geneva
Association)
Insurability: Limits and Issues
The limits to insurability are both of a formal actuarial nature and a politico- legal
character. The diagram below shows the severity and frequency of losses of different
events, which is one aspect of the insurability or not of different types of risks. The
relationship between the severity of an event and its frequency is shown below. This
diagram allows us to draw a theoretical boundary below which risks tend to be insurable.
Nuclear
High
power
Terrorism
Severity
Insurable risk
Car Ins.
Life Ins.
Low
High Frequency Lo w
Insurance involves pooling similar or comparable terminal risks, to spread losses over
time. Terminal risks are risks with high frequencies and established losses, and are
therefore insurable. Risks with low frequency and a high loss potential, such as nuclear
power, are not insurable. Basically, insurable risks are the ones below the diagonal. For
the insurance industry, loss prevention and risk management are important tools as they
help move risks from above the diagonal into the insurable realm, thus creating more
business opportunities.
The diagram only displays the limits to insurability from frequency or severity of loss,
and there are several other insurability limits. There must be a maximum loss figure that
10
can be put on a risk, which is why an act of terrorism or war cannot presently be insured.
Another limit of major importance for insurance companies is that the people involved
must be prepared to pay a premium. If the insured do not see the need for insurance at
the existing terms, then there is a need for mandatory state legislation making people take
out insurance. Car insurance is a good example where many states have made it
compulsory, as people tend to overestimate their own driving abilities. A recent survey in
Scandinavia asking people about their quality of driving found that 80% of people
regarded themselves as above average drivers, whereas clearly only 50% could be above
average drivers. Another insurance principle is that both insurers and insured should be
worse off after a loss. If the insurer or the insured makes a profit out of a loss then
something about the idea of insurance is not done right.
In the last few years it has become very popular to say that banks and insurance
companies are the same thing. However, banks manage assets, insurance companies
manage liabilities. I think we all know that assets and liabilities are not exactly the same
thing. They need a different way of management. In other words, insurers are experts in
dealing with uncertainty by asking for a premium today for something that may happen in
20 years time. Costing risks is the big know-how of insurance companies.
With regard to the red, or medical, life sciences, risks are insurable if they are limit ed to
one generation, or in other words, if the risk that you are covering will die with the
patient. If the risk remains after the patient is gone, in another generation, then it is not
an insurable risk as you cannot put a loss factor on it. There is exactly the same problem
with the green life sciences, such as GM crops. There is no reason why you should not
be able to insure GM crops if they have terminator seeds, but these seeds are not very
popular with politicians or farmers, as they deprive that part of the farmers‟ independence
which comes from being able to use part of the previous year‟s crops for planting the
current year crop. But on the other hand, without terminator seeds the product becomes
uninsurable, as it becomes impossible to put a time limit and therefore a loss limit to the
associated risks.
Contextualising Risk: Opportunity and Vulnerability
It is important to point out that one of the big mistakes, which is especially apparent in
the debate on life sciences, when discussing risks, risk assessment and perception, is the
failure to include a discussion of opportunities. There is a need to balance risks and
opportunities. Americans tend to exaggerate opportunities. Europeans tend to exaggerate
risks. Risks are only relevant if they concern people. No one would be willing to pay a
premium to insure against the risk of an avalanche somewhere at the North Pole, as there
are no people involved or affected. Risks are context and situation dependent: the same
event in a different location or at a different time is a completely different risk. The link
is always the link to people or to economics.
The last 10 years of risk management were simple, as there was agreement in insurance
that the biggest risk, the first 100 million dollar loss, would be an earthquake in Tokyo or
a similar event. September 11 has told us that society‟s largest risks do not come from
11
nature but from people, and that it is no longer enough to consider the maximum possible
risk but instead the maximum possible imaginable risk. If anything, September 11 has
shown that anything imaginable might actually happen, which changes the upper limit of
the diagram. For example, take the hypothesis of the fourth hijacked aircraft in the US
that, according to many insiders, was aiming for a nuclear power station. If you crash a
Jumbo jet into a nuclear power station, it will most probably not resist the impact, which
would give you the risk that so far nobody in the whole electricity or energy industry has
ever even calculated. It could be 70 times Chernobyl. September 11 has led to a
complete rethink of what a maximum loss could be.
Insurance and Sustainability
Insurability can also be a means of influencing the choice of technology, as it puts a cost
on future risk. Therefore, with mandatory and unlimited insurance, insurance used as a
free market instrument would include the future cost of risk of different technologies,
thus providing clear market signals as to more or less risky, or sustainable, choices of
technologies. For example, nuclear risks were not insurable, which is why states have
defined limits of liability; nuclear power stations are only liable for the first 200 million
dollars and anything above that is defined as an Act of God. We should be careful not to
make the same mistake when choosing new technologies.
Loss prevention and ethics are basically sustainability concerns. It has been estimated
that 80% of medical costs today could be avoided if people did not smoke or drink and
had reasonable time to exercise. Nobody seems to be interested in this, and insurance
does not really bother because as long as people are prepared to pay the increasingly
higher premiums there is no problem for insurance. But at some point you could reach a
limit of insurability, if people no longer can afford to pay premiums. Then there is the
problem of getting the state to join in or people will, as in the US, live without health
insurance. In both cases you have a social mess, but the alternative would be risk
management to bring down the losses to a level at which people can pay a premium.
In the Netherlands, as a result of the big flood in 1962, flood insurance has been declared
illegal, making it very clear that only engineers can prevent the Netherlands getting
flooded. Insurance does not avoid flooding. Insurance can only compensate part of the
loss, and only the financial loss. We must be very clear of the role of insurance here.
Insurance cannot recreate things that are lost.
Very often there are much more efficient ways than insurance to prevent losses, such as
good engineering. Excluding car tail collisions from insurance, as has been done in
California, has enabled premiums to be reduced by 25%. So there are some accidents or
risks where insurance can really only cover the damage, but there are other measures
much more efficient in preventing the potential risk through changing behaviour.
Some people think that the insurance sector is risk averse because there are many
uninsurable risks. If the insurance sector were risk averse you couldn't buy life insurance
because the only certainty of life is death. At least you can be sure of this because if you
12
look at the diagram, you have a very high frequency because everybody will die, and
very low potential maximum loss. This brings us to the actuarial sciences where it is
possible to calculate the price on human life. If you have read The Sceptical
Environmentalist, you know that these values in risk management and loss prevention are
very variable. A $100 investment in a seat belt saves a human life. With asbestos, the
equivalent figure quoted was perhaps 50 million dollars. Even without considering
sustainable development, this raises many technical questions. However, insurance
should now be regarded as a tool of politics and society, rather than just an abstract
economic principle.
Conclusion
In the past and for many economists today, insurance is considered a luxury. In other
words when times were bad the money spent on insurance went down, and when people
had too much money they had more insurance. If you look at the statistics of the last 20
years you will see that insurance goes up 5% in good times and down 6% in bad times.
In other words, insurance today is a pre-condition for both companies and technologies to
exist. Insurance has become a pre-condition for technologies, for economic development
and in that sense I think it is very interesting what we are trying to do with this
INTEREST project: to understand what role, although this role will always be limited by
several factors, insurance could play, in terms of calculating risks, of helping to choose
the best technology for sustainable development.
Discussion
It was noted that 20 years ago, terrorism was insured without premiums. Drawing
boundaries between political, natural and technological risks may lead to the insurance
sector starting to exclude one or more of these types of risk, and there is a tendency to
transfer risks rather than trying to reduce them. Insurers and society need to engage in a
dialogue in order to define the role of insurance in this context. Should we focus on loss
prevention or risk prevention? However, it was also pointed out that September 11 was
not a typical event and we should be careful in drawing too many conclusions from it, in
everyday life industrial risks for instance continue to be more important than this singular
event. The need for INTEREST to focus on the precautionary principle and insurance
was stressed.
Moving on to the subject of innovation, the important role of social psychology was
raised: people do not behave as expected, and often technologies are created to make
them behave in a certain way. This topic was dealt with more formally in the presentation
on innovation (see paper below). In innovation, which is dynamic and not static, things
change by definition, with important implications for insurance. Asbestos for instance
was not insured as the risks were not imagined, but there is a need to investigate future
scenarios even for unimagined events. It is questionable how realistic the prospects are of
13
this occurring in practice, as it was also pointed out that a lot of risks were noticed but
subsequently ignored, such as BSE. Insurance and innovation do not always go together.
However, insurers have companies to run, and cannot cover all risks. A financial
assessment should include social, political, economic, technological and legislative
assessments. Also, investments need to be made in order to finance the insurance of risks.
It is known that the investment and underwriting aspects of insurance tend to be
disconnected. Some insurers hold the view that risk assessments are a luxury and will
always be offset by investments, but events in 2002 – September 11 and the crash of the
stock exchange – point to the opposite.
The ethics aspect was briefly mentioned, noting that high risk groups will remain to exist
and one part of the insured population will pay for the other part.
Finally, the position of the INTEREST project in relation to the EU 6 th Framework
Programme was discussed. The 6FP centres around the creation of networks of
excellence and integrated projects, with improved emphasis on foresight and on including
EU candidate countries. The INTEREST project should consider a proposal topic under
this framework in time for the final workshop.
14
2.2 When is a Risk Insurable?
Gerry Dickinson
City University Business School
Introduction
Insurance is a legal contract whereby an approved party, the insurer, in return for a
payment, undertakes to pay another party, the insured, a sum of money (or an equivalent
in kind) upon the occurrence of a specified event(s) which causes a financial loss to the
insured. Insurance contracts provide financial indemnification for losses arising from a
defined set of risk causes, mainly from acts of nature, from human error and malfeasance
and from other accidental causes.
Providing these contracts of financial indemnity is the core business of insurance
companies. However, it is in their self- interest to help minimize the loss of physical
assets and damage to the environment and to maximize human safety. This interest in
“loss prevention” makes the principle of sustainable development relevant to the
insurance sector. Nevertheless, the insurance sector must operate in partnership with
others, e.g. government and other sectors.
Insurance markets have often been based on “mutuality”. Companies operating in the
same sector often share risks. Property and casualty insurance tend not to be mutualised.
Mutual companies are in a position to cross-subsidise between different types of
consumer when providing life insurance for example.
The case studies covered by the current workshop refer to different types of insurance: a)
property insurance (NYMPHEA); and b) liability insurance (GM crops).
Limits of Insurability
There are clear constraints on the types of risks that can be insured and on the magnitude
of these risks that can be transferred to insurance markets. These go beyond simple
considerations of the expected severity and frequency of losses, though these are in
themselves key factors.
The capacity of insurers to provide insurance is constrained by: a) regulatory and legal
limitations; b) pricing issues; and c) the capacity to provide sufficient risk transfer. There
are also constraints on the demand for insurance products.
Regulatory and Legal Limitations
Insurance legislation varies from one country to another. Legislation will specify what an
insurance company can supply under its license. Hence, what is meant by insurance in
any particular country is effectively determined by the State through the legal system.
15
New insurance products may be restricted because they fall outside the list of permitted
insurances. Certain types of insurance, for example flood insurance in the Netherlands,
may be deemed to be against the public interest. EU Directives in the mid-1990s unified
the list of types of insurance that could be supplied. However, certain new methods of
securitisation are not on that list.
In addition, insurance contracts must be legally enforceable. This usually requires that:
(a) the insured party suffers a financial loss (the principle of insurable interest); and (b)
the insured party does not profit if the agreed event(s) causing the loss occurs (principle
of indemnity).
Pricing insurance
Insurance companies must be able to price their products in an economically sustainable
way. There are certain classic criteria for determining whether or not a price can be
established and the risk rendered insurable.
First, the events causing or triggering insured losses must be sufficiently well-defined.
Second, an insurer must have information to be able to estimate with a sufficient degree
of accuracy the likelihood and severity of losses from the insured events. This also
presupposes that there is no significant adverse selection by potential insured parties, i.e.
that those most likely to suffer losses do not disproportionately seek insurance. Third, the
existence of insurance protection should not cause the behaviour o f insured parties to
change the frequency and severity of insured loss in an unpredictable way. In other
words, moral hazard should not reach unacceptable levels.
Since private insurance companies mainly operate in competitive markets, they must
charge prices that reflect their expected costs, which include both payments in respect of
losses and other administrative costs. If the supply of insurance is to be sustainable over
time, there cannot be too much cross-subsidisation between different types of consumer.
If companies charge consumers prices that are higher than the cost of indemnifying
losses, consumers will switch to another insurance company. If companies charge prices
that are too low, they will be unprofitable. Shareholders will not then be fo rthcoming
with capital to absorb the risks.
Capacity to Provide Sufficient Risk Transfer
The capacity of the industry to provide risk transfer can be defined as the financial
resources of the insurance market in relation to the scale of potential losses. The financial
resources of the insurance market fall into three categories: a) the capital and reserves
held by insurance companies and their new capital raising capability; b) part of the short-
term cash flow from new business, since the insurance compa ny must be viewed as a
„going-concern‟ and c) the capital and reserves held by the global reinsurance network
and its capital raising capability. The importance of the short-term cash flow factor is
illustrated by the fact that, after a very large loss, insurance prices tend to rise.
16
The capacity of the insurance industry also depends on the efficiency of its risk
management support, since the scale of potential losses depends on preventative action.
This includes: a) the efficiency of loss prevention and safety management support
provided ex ante by insurance companies and others; and b) the efficiency of its post loss
event activities, e.g. early remedial action and support, and a speedy, flexible approach to
claims.
Figure 1 maps the risk transfer network associated with the insurance industry.
Individuals, organisations and governments may place insurance contracts with direct
insurance companies who in turn can transfer risk through reinsurance contracts. In effect
insurers access the capital of reinsurers when they purchase reinsurance. Each of these
groups may establish derivative contracts with investment banks and the financial
markets in order to hedge financial risks. Both insurers and reinsurers may securitise their
risks by placing them directly onto global capital markets.
Figure 1: The risk transfer network
Individuals,
organisations
and governments
Insurance contracts to
transfer insurable risks Derivative contracts to hedge
financial risks, eg currency risks,
Direct olacing of Direct insurance interest rate risks, commodity
risks onto global price risks, weather risks etc
capital markets companies
(ie risk Insurance derivatives
securitisation)
New capital Reinsurance contract
raising transfers
Reinsurance Insurance derivatives Investment banks
companies and financial
markets
New capital
raising
New capital
Capital markets raising
(global)
17
Figure 2 shows how the risk transfer market tends to handle the spectrum of risks defined
in terms of frequency and severity. Most people and organisations would tend to self-
insure low frequency/low severity events when they have the financial capacity to absorb
losses. Higher frequency, low- medium severity risks tend to be covered by national
insurance and reinsurance markets. Higher severity risks will be covered by global
insurance markets and national or regional reinsurance pools. The highest severity risks
must be securitised on to global capital markets or may be covered by government
finance or guarantees.
Figure 2: Spectrum of Financing of Catastrophic Risk Exposures
insurance and reinsurance
risk retention on national markets
(self-insurance)
international reinsurance and
loss frequency
national/regional reinsurance pools
risk securit isation onto global capital
markets and/or government (inter-
governmental) finance or guarantees
low high
loss severity
Global Capital Markets and the Role of Government
Insurance/reinsurance markets have the capacity to absorb only a limited number of
catastrophic losses over a short to medium time period. The insurance industry‟s free
reserves are of the order of $150 billion, three times the level of claims arising from the
September 11 attacks. In all, the capital and free reserves of the global non- life insurance
and reinsurance industry are currently about $550 billion. Risk securitisation of insurable
risks on to global capital markets can provide greater capacity. The size of global capital
markets (equity and bond markets) is about $50 trillion and so a daily fluctuation of 1%
in their value is $500 billion. This is much larger than most conceivable insurable losses.
Thus global capital markets have more potential capacity to absorb risk than any national
government.
However, risk securitisation requires the precise pricing of risk in relation to its frequency
and severity. The precise pricing of such risks in advance is difficult. In addition there is
limited scope for cross-subsidisation in capital market transactions, either over time or
between classes of consumer.
18
Any government guarantee and/or capital support comes ultimately from present or
future taxation. Hence it will be conditioned by other economic and political trade-offs.
However, government guarantees/capital support can be more flexible since they can take
into account the broader public interest and can have a longer time perspective. There is
more scope for cross-subsidization and a higher level of moral hazard can be accepted.
For particular causes of loss, such as terrorism, and indeed war, governments should play
a larger role, since they are in a better position to prevent or reduce these risks than the
private sector. Moreover, if they are paying they also have the incentive to reduce the
risk.
Demand factors
So far, this paper has addressed supply factors affecting the insurability of risks.
However, demand factors are also important. If prices are set too high, this will make
insurance unaffordable or, more commonly, consumers will perceive it to be
unaffordable.
There are also key policy issues associated with natural catastrophes such as floods,
earthquakes and windstorms. These events tend to impact more on certain
individuals/organisations than others. If insurance companies charge fair prices in
competitive markets, it means that those most affected will face high, and possibly
unaffordable, prices. This is an issue of cross-subsidisation. The industry believes that
only the government can effectively provide a compensating subsidy. This can be seen as
falling within the remit of government‟s wider social and economic policies, especially
policies on the redistribution of income and wealth.
Discussion
Even if governments do get involved in providing insurance of last resort, it was
observed that the insurance sector might still be needed to implement and administer the
system. In expanding the pool of insurable risks, there is a need to get insurance products
down to the grass roots. Some insurance companies have priced themselves out of small
risks – the example of liability insurance for consultants was cited. Government and
insurance could benefit from working together on these issues.
The pricing of risks was identified as an important issue. Sustainable development related
risks are difficult to price. For example, the level of risks associated with pollution might
vary because governments could, unpredictably, re-define the event that triggers a
payment. In practice, the insurance sector is inclined to be flexible about whether an
insurable event has actually taken place, leaning to the generous side. Corporate
reputation could be at stake in this type of issue.
19
The discussion showed that there were social and educational aspects to insurance. If
someone faces catastrophic risks they ought to be insured. However, small relatively
inconsequential risks should not be insured. The top and the bottom of the market - big
and small risks – were particularly important in emerging markets.
The issues of moral hazard and special interests were raised. The reason that third party
liability insurance has to be compulsory is that the risks are borne by parties other than
those taking out the insurance. The irony in the case of nuclear insurance is that the
insurable liabilities associated with production assets exceed the insurance liabilities to
third parties as a result of international agreements!
Finally, it was observed that people‟s decision- making processes matter, though markets
are frequently not transparent in any case.
20
2.3 Innovation and Insurance: Preliminary Investigations
Bastiaan de Laat 1
Technopolis
Introduction
This paper discusses recent insights from innovation theory that seem important in order
to reflect on the relation between insurance and (technological) innovation. The paper
will discuss the non- linearity and non-causality in innovation, path-dependency and
innovation being a collective and interactive process between actors (heterogeneous
networks). Next it will address the difference between “risk” and “uncertainty,” an
important distinction in order to understand the pote ntial role of insurance in
technological innovation. Finally the paper points out some opportunities and possible
directions to go from here. The main argument of this paper is that in order to think about
insurability within innovation processes, one may have to give up the very idea of
calculability of risk, and engage in a process to involve relevant actors.
Three lessons from innovation studies2
Developing tools for insurance in research and technology policy requires a discussion of
the findings from technical innovation studies. Even if general theories regarding
research and innovation differ enormously between, for example, sociology and
economics, lines of consensus can be identified which, in our view, may have
implications for thinking over developments in insurance.
This section discusses three major results from innovation studies and reviews their
implications for futures thinking. The first relates to the observation that innovation is not
a linear process from science to market. On the contrary, it consists of the gradual
establishment of links between actors. Second, innovation processes consist of the
formation of heterogeneous assemblies, being simultaneously social as well as technical.
Last, innovations are path-dependent and can quickly become (quasi-) irreversible.
Innovation as an interactive process
Innovation has long been described as a stepwise process from science to the market.
Schumpeter (1934) was the first to analyse technical innovation processes more
thoroughly and to conceptualise technical innovation in stages of invention, innovation
and diffusion. During invention, new ideas (sketches, models) are generated. In
innovation some of those inventions can be picked up by firms and brought onto the
market. Finally, the diffusion phase consists of imitations of the original „prototype‟
adopted by a larger and larger number of adopters. The reason for a technology not being
adopted, even when profitable, would lie in the conservatism of the group of potential
1
I gratefully acknowledge my colleague Soraya Fah my with her assistance in the literature review.
2
Earlier and more extended versions of this discussion can be found in De Laat 2001, De Laat and
Larédo, 1998 and De Laat, 1996.
21
adopters. In sum, as in a relay race, actors in the innovation process are supposed to
deliver their products sequentially – first knowledge then a concept or patent, then a
prototype, etc. As in the race, each withdraws having delivered their finished
contribution. This process was supposed to be governed by a logic of either science-push
or demand-pull. 3
However, empirical studies show that innovations do not follow such a sequence. Rather,
the objects developed and elaborated by actors describe a whirlwind pattern. Products,
whether they are consumer products or scientific articles, are developed and tested,
maybe normalised, but will also encounter resistance, be reshuffled, redefined, tested
again, and so on. They can go from science to market, but they may also find their origin
in a certain demand. In other words, scientists sometimes play a dominant role, but as
Von Hippel (1988) showed, users can also be important in this process. Finally, invention
can even take place in what normally is called the „diffusion‟ stage (Bijker 1992). No one
general factor, no one single social group can solely be held responsible for the
development of an innovation.
Innovation studies suggest that innovation is not a process whereby actors (laboratories,
firms, users…) intervene sequentially, but one during which durable links are bit by bit
created between these various players. As stated earlier, foresight exercises themselves,
being embedded in innovation processes, are typically of such a collective nature.
Symmetrically, futures thinking may want to account for the collective character of
innovation.
The social and the technical are simultaneously shaped
The second insight derived from innovation studies concerns the relation between the
technical and the social. The technical, and more generally the construction of technical
systems, was until recently the privileged domain of the technologist, the scientist or the
engineer. The social world and within that the study of interactions between human
beings was the realm of the sociologist, the behavioural psychologist or other students of
human behaviour. Micro-studies of innovation have however shown that the within
technical innovation the social and the technical are intimately linked: technology is
shaped by human beings – thus within social groups – but, once shaped, technology
implicitly or explicitly defines the social environment in which it is able to function: a
group of researchers can work with the most sophisticated scientific methods on the
technical performance of a Lithium battery for the electric vehicle, but it also implicitly
makes the hypotheses that a world can be built in which the vehicle that will incorporate
their battery will be able to run.
Many authors have demonstrated that the social and technical are shaped in one
movement: whether we look at bicycles or bakelite at the beginning of this century, a
gasifier in a developing country or high- tech subway and aircraft projects, 4 in all cases
we see that the social and the technical are shaped together. Techniques are shaped by
3
Coo mbs et al., 1987 g ive an extensive overview of the technology push -demand pull debate.
4
See respectively, Akrich, 1993; Bijker, 1987; Latour, 1993; Law & Callon, 1992.
22
collective social action, simultaneously shaping the politics, the users, and the
infrastructure associated with them (Callon, 1987). And once established, techniques act
in themselves (Latour 1994). Hence, a technical object does not only reflect social
relationships (Bijker, 1995) but also “…transcribes and displaces the contradictory
interests of people and things…” (Latour 1992). Therefore, innovation – as a process as
well as in terms of its outcomes – is a socio-technical assembly in which multiple
heterogeneous actors play a role. For the purpose of our enquiry it means that research
priorities and technological options are always connected to certain configurations of
actors – those through which they are constructed as well as those which are projected
onto the future. Put another way, every technical choice implies or presupposes a social
configuration, and vice-versa.
Irreversibilisation occurs … and often rather early
The third implication of innovation studies for insurance thinking stems mainly from the
work of economists of technical change and their observation that innovations are path-
dependent. Authors amongst which the most well know are Dosi, Nelson and Winter
describe technological innovations in terms of trajectories, accounting for the existence of
paradigms within technical innovation that are relatively fixed over long periods of time. 5
These paradigms are built up little by little, resulting in the end in shared frameworks. As
David showed for the QWERTY keyboard, such trajectories are often far from „rational‟
or „optimal‟ with regard to the technology under concern. They heavily depend on initial
choices becoming (quasi-)irreversible.
David (1986) exp lains that in the case of QWERTY, the arrangement of the keyboa rd was the most optimal
way to prevent the compacting of typebars one behind another. For the manual act of typing itself more
optimal solutions might have existed (by 1867 there were more than 50 patent attempts describing a
commercial typewriter). Moreover, Remington was near to broke when it bought the manufacturing rights
fro m the inventor. Nevertheless, schools for novice typists based themselves on the Remington machines.
Likewise the early manuals. Consequently the supply -line of operators were trained in QWERTY,
influencing buyers and suppliers. QWERTY's final standardisation in 1890 came fro m the expectations that
the buyers of typewriting machines had with regard to the equipment users, who were now used to the lay -
out. Despite the fact that in 1890 new machines had been developed with both a better key arrangement and
a better way of preventing the typebars from co mpacting it was already too late: now, some 100 years later,
the greater part of the world still finds itself typing on a QW ERTY keyboard. For the other parts a
symmetrical story can certainly be told (France for instance uses 'AZERTY' -type keyboards). Note that
observations of the sociologists cited in the preceding paragraph and the economists cited here converge:
the QWERTY-story shows very well that the simultaneous stabilisation of both social and technical
elements accounts for irreversibilisation and path dependency of innovations.
Other authors argue that stabilisation of technological trajectories occurs in early stages,
fostering the development of a certain trajectory and excluding others – however this
often passes unnoticed, and comes out only once the trajectory indeed is stabilised. This
was already pointed out in the so-called “Collingridge” dilemma which stated that,
around a given innovation, in the course of time it becomes more and more difficult to
influence technical development whereas at the same time it becomes more and more
visible and open to social influence (see exhibit).
5
Dosi, 1982; Nelson & Winter, 1977; Sahal, 1985.
23
Exhibit: Collingridge Dilemma
Technical choice
Social choice
Collingridge’s dilemma as a way to express irreversibility of
technological trajectories: the number of technical options
decreases whilst the will to have “social” influence on that
development increases
Indeed, small but critical events appear important in any technical development, whether
this is within research programmes (Rip 1995), in the case of products competing for a
market share (Arthur 1989) or for technological systems (Cowan 1990). Hence the third
consequence of innovation studies for insurance thinking is the need to identify those
areas where early closure of socio-technical choices is at stake. Methods such as
regression analyses, logistic – „S‟ – curves, envelope curves, or other types of
extrapolation can indeed be used to describe trajectories but they eventually will have
predictive power only after irreversibilities have been created.
The management of risks associated with technological innovation
What are current challenges for insurance posed by scientific and technical
developments? This section briefly reviews the increasing complexity of risk, the shift
from risk to the concept of uncertainty and the corresponding distinction between
prevention and precaution.
The complexification of risk
Insurance is normally based on an assessment of the risk that a certain event may occur.
Preferably this is done through a calculation of the probability that the event happens.
Thus, living in an area with a high rate of burglaries will increase the premium of your
insurance policy against burglary. In several countries women drivers pay less to insure
their car since statistically it is shown that they cause less losses to insurers.
24
As long as probabilities can be calculated, a premium can normally be set. However, as
recently underlined by D. Kessler in the French report on insurance 2000 of the FFSA,
the universe of risks is expanding. In parallel to the traditional risks, often natural ones,
new risks emerge due to scientific and technical developments, economical and financial
evolutions and social transformations. Not only the quantity, also the nature of risk is
changing. Risks become more progressive than accidental, they are more sustainable
than in the past and sometimes irreversible as in the case of environmental damage.
On the one hand, it may be suggested that scientific and technological progress would
make risks more foreseeable. For instance, meteorological modelling and prediction help
to calculate probabilities for harvest losses and may thus assist in calculating a premium
for an insurance policy for a farmer. However, scientific and technological progress also
bring along new elements that were not known before, make risks more and more
interdependent and one event can be the source of a chain of consequences that as yet
may be unknown. This evolution of the nature of risk implies an adaptation of
management tools – or even of the notion of risk itself.
From risk to uncertainty
As argued by Sinclair-Desgagné and Vachon (1999), managing technological risk means
being able to assess risks and, once assessed, to share them. These authors see insurance
as one of the four ways in which today risk is dealt with, the other three being
assessment, prevention and mitigation. Two fundamenta l dimensions have, according to
these authors, to be considered in order to assess risk: the magnitude of the potential
outcomes and the probability attached to them. Three approaches are used in risk
assessment
qualitative and quantitative methods
risk perception
the acceptable level of risk
Now, if one wants to insure risk associated with the potential consequences of a given
technological innovation, one should be able to measure and calculate this risk. What
does this mean in practice? In practice this means that for any given technological option6
one would have to be able to calculate the probability of occurrence of all its potential
negative effects, i.e. those that would potentially involve a loss for the insurer, could be
calculated. It is clear that this cannot be achieved. The following example may draw out
two main reasons for this inability of insurance to deal with innovation.
Let us go back to the pre- mobile phone era. Twenty- five years ago it would have been
difficult to imagine that, in the future, one out of two European citizens would have a
cellular phone. Today no one would object to this hypothesis, which in the past would
6
For the sake of the argu ment let us simp ly assume that a “technological option” is an “invention”
with a certain technical content, which may or may not be realised in the future.
25
have sounded futuristic. 7 It has become part of a West-European city scene to see the
streets, trains, bars, schools, etc populated with people showing the today very familiar
picture of “the hand to the ear.” It sometimes even looks as if they are talking to
themselves – but in fact they use an earplug... 8
Twenty- five years ago a car phone was a real luxury, and the impact of this GSM
technology on our daily lives, as well as on our everyday behaviour, was not only
“unknown,” it was simply hardly imaginable. And even less people would have imagined
that, based on microwave technology, a cellular phone may lead to up to 2 degrees
temperature increase in the part of the brain directly surrounding the ear. Experiments
show that the diameter of the heated zone can be up to 2 cm. Nobody really knows what
the consequences on brain tissue will be, and whether any sustained brain damage may
arise from this effect. Science is still in full controversy about the real consequences,
although telephone operators have started to think over the question and even launch
“mobile phone and health” programmes. But of course, this was far from being the case
twenty-five, ten, or even five years ago.
This simple example shows that in dealing with innovation, the insurer may be
confronted to two interrelated problems involving controversies around the negative
effects (on health, on environment…) of new technological developments. The first is
simply that in order to only imagine potential effects, one should be able to think about
the possible states of the future world (or the possible future worlds) associated to the
innovation at stake. For instance, which part of the population will use the cellular phone,
for what purposes, how does this potentially change behaviour, … In other words, what is
the “socio-technical script”9 of the object or the object that is promoted through the
innovation. If the “innovation” only exists on paper, as is often the case, imagining such
future worlds is a difficult job, and few tools have been designed to think about the issue.
The second problem is that, even if the technology matures and is adopted, a nd even if at
a given moment a physical effect is objectively established, scientific controversy about
the effects of such effects will unavoidably arise. Even if one may objectively measure the
temperature increase in the brain, the effect on the brain o f this temperature rise is highly
controversial.
In a recent book Callon et al. (2001) 10 propose a similar definition of risk as the one cited
earlier. Risk is a well- identified danger associated with an event or series of events which
can be perfectly described. One does not know whether they will occur but we know
that they might happen. In the case of technological innovation however, involving
potential negative effects surrounded by scientific controversy, one is no longer
confronted to risk but to uncertainty, which is something different. According to Callon
et al. (p.40, our translation), uncertainty means that:
7
The today very serious clothing designs, which integrate electronics would have looked like p ure
science-fiction a decade ago.
8
Though in writ ing this we realise that it is still a somewhat strange picture if you come to think of it.
9
Akrich, op.cit.
10
26
You know that you don‟t know it, but that is about all you know.
In other words, a situation of uncertainty is the situation in which scie nce is not capable
of describing all possible worlds.
The natural reaction of an insurer will be to ask at which point of the stabilisation of the
technological innovation, and at which point of the closure of the controversy, he may
decide to turn an uncertainty into a risk. Another natural reaction will be to continue
measuring the phenomena as they occur, and try to derive regularities, and not
necessarily causalities. An insurer does not necessarily feel the need to know whether the
increasing number of storms in a given region is due to global warming or to something
else. In fact, it would be enough for him to know the increase in natural storms in that
region, in order to calculate the future probability with which storms may occur in order
to fix the premium, for the insured inhabitants, for losses caused by storms in that region.
The latter position would however imply that the insurer would engage in the discussion
only when both innovation and controversy are to a great extent stabilised. The objective
of the INTEREST project instead is to upgrade the reflection on the issue, and to consider
the role of the insurer more upstream in the process of technological innovation and
scientific controversy.
From prevention to precaution
The present one is not the first reflection on technological innovation, risk, and insurance.
The last decades have seen several industrial disasters (Minamata, Bhopal, Seveso, etc)
which fortunately have allowed progress to be made on the matter. Even if the attribution
of responsibility, especially in case of disperse environmental pollution, is not always
easy to establish, a societal process of prevention of such risks has been set in train since
the 1970s in several industrial sectors, and several directives (the best-known of these the
Seveso directive) have been conceived of in order to decrease environmental risk and
pollution. Nevertheless, disasters still occur often, even in Europe, as again very recently
experienced with the explosion of the fireworks factory in Twente or the chemical plant
in Toulouse.
Although these risks are strongly linked to industrialisation of society and thus are often
typically viewed as technological risks, the potential role (not always fulfilled besides) of
the insurance sector is clear. Whereas Sinclair-Desgagné and Vachon (1999) suggest that
insurance is an instrument parallel to prevention, we suggest that insurance can be very
important as a regulator and has a role to promote prevention. The reason is that, for an
insurance company, pollution prevention most of the time equals loss prevention. Even if
this role is known, insurance companies know how to go about it, given the fact that
industrial risks are apparently still great, insurance‟s role may be reviewed and eventually
reinforced. In the words of Sinclair-Desgagné and Vachon
By design, the terms of an insurance policy, namely the premium, the
extent of coverage, restrictions, exclusions and deductibles, should
27
separate and deal with the various sizes and types of risks involved.
Insurance contracts then become a mechanism that provides incentives to
reduce risks. The insurer thereby takes the role of a surrogate regulator.
Prevention is something that insurance can be an incentive for. There is a second type of
technology related risk in which the role of insurance seems less well defined – and is
most probably less well definable. It is related to the concept of uncertainty introduced
earlier. Let us take again the example related to climate change. It is generally
acknowledged that the climate is changing. In particular, the number and intensity of, for
example, storms and drought around the world shows a changing pattern. For the insurer
however, risk often cannot, or not yet, be calculated, but meteorological services become
an increasingly important ally for insurance companies. Insurance companies may protect
themselves by requiring high premiums for people living in “risky” areas, or living in
housing which would not be protected well enough against potential calamities. In
practice however one also sees the emergence of “uninsurable” zones in which the State
takes over the role of insurance company since losses become too high.
At what point in time can one possibly start to imagine the effects associated with an
innovation or with an industrial activity? Today, the hole in the ozone layer is generally
attributed to be due to forms of human industrial activity. The first is, amongst others,
attributed to the presence of CFCs in the atmosphere, used in a great variety of industrial
processes and consumer products. Establishing this causal link however has required
several years of controversy before the scientific facts were “stabilised.” The first
observations of a hole in the ozone layer date from several decades back but were not
instantly followed by a general acknowledgement of the existence of an effective ozone
depletion. As a matter of fact, initially the measurements were not believed by the
scientists involved since, outside of the theoretical measurement range, they would be
“impossible to occur.” This appeared a wrong assumption. As concerns global warming,
the first theories on the influence of industrial activity on a potential greenhouse effect
date back to the 19th century, although the real effects could only be guessed and were not
clear. It often takes a great effort to transform theories – especially when there are several
competing ones – into practice, especially when moreover they are strongly connected to
politics. 11
In order to respond to the situation of uncertainty, the precautionary principle was coined
in the beginning of the 1980s especially in the environmental area. It indicates that it is
justified to adopt measures at a reasonable cost in order to prevent major losses even if
scientific knowledge is as yet not certain about whether they will ever occur or not. In the
case of prevention potential risk is known and should be minimised by well-directed
action, in the case of precaution even the risks are not yet known but may eventually be
great. In other words, precaution may lead to a decision to not decide (to go ahead on a
given technological trajectory for instance), awaiting more certain information on the
risks involved. Even if that may sometimes be the case (and in fact it is an interpretat ion
generally given to the precautionary principle, Callon et al. (ch. 6), indicate that the
11
Note that our argument is not to disconnect the two, rather the contrary (see B. Latour, 2000,
Politiques de la nature, Paris: La Découverte).
28
principle should not paralyse the debate (acting only at zero risk) in order to, in fine,
avoid discussions on responsibility. On the contrary, it should serve to organise debate on
technological risks and options by allowing the different actors involved to design the
procedures with which they will construct the debate.
So even if both can be linked to technological risks, there exists a clear difference
between prevention and precaution. One can apply the idea of prevention if potential
risks are known and measures can be taken as to minimise them as much as possible.
Prevention, in our view, is applicable to some recent cases of industrial catastrophes. In
such cases, insurance can play an important role as regulator and controller of risk by
promoting prevention. Just like in order to be insured in a burglary-sensitive region one
may be asked by its insurance company to put an extra lock on the door, specific
measures can be asked to be taken for certain risk sensitive industrial companies.
Although in some countries/sectors insurance companies play such a role, it does not yet
seem to be general practice. 12
Precaution has to be dealt with differently. In the las t section of this paper, we outline
some opportunities to place insurance more upstream in the discussions on technological
innovation and scientific controversy.
Opportunities
The complexity of technological risk has today grown to such an extent that it can easily
be admitted that such risks ask for specific competence, if not specific theorising on how
to be dealt with. Traditional insurance policies and contracts seem not adapted because of
their restrictions, their maximum guarantees and the time factor. Insuring big natural
disasters – but also “smaller” collective losses caused by specific products – can in the
end only be re-reinsured by the public authorities. However, it would be more efficient to
collectively manage and anticipate them. This last section sketches some opportunities in
order to do so.
The idea of irreversibility outlined in the first section implies that a new scientific or
technological trajectory wins because it is chosen, it is not chosen because it will win…:
a major problem in technological innovation is that technical choices are made, not
always on “scientifically rational” grounds, and they are harder and harder to become
undone the more their environment is shaped to fit them and vice versa. In other words,
one has to act early in the process in order to have an effective influence on the direction
of a trajectory. We propose that (at least) three opportunities exist to deal with the
management of innovation for insurance purposes, in a situation where risk cannot be
easily assessed and uncertainty is the rule. These are the following
Watchdogging13
Engaging in the R&D processes
12
In a large survey on the incentives for chemical co mpanies to reduce environmental risk, insurance
companies indeed came out as one of the potential drivers though not yet fully aware of this ro le.
See Schot, et al, 1991.
13
This is referred to as „horizon scanning‟ in the UK.
29
Engaging in the decision-making process
These will be briefly discussed in turn.
Watch-dogging
For the insurance sector (and more generally for technology management) watch-dogging
means to create an “early warning” function which may anticipate new future
technological trends, new scientific controversies, and, at the end of the day, new
potential losses for the insurance sector. Strangely enough, this function seems not very
well developed in the insurance world. Within one of the major global reinsurers, we
encountered only one person whose role it is to keep an eye on new unexpected events
and perform a form of worldwide “science”, transforming into “claim”, watch with a time
horizon of over 5 years.
The basics of such a watchdog function can be sketched as follows. The simplest version
of the watchdog function is to monitor, first of all, newly emerging types of claims. 14
Such a screening may result in, for instance, the observation that in certain regions of the
world some new, as yet inexperienced insurance claims are occurring. The idea then is to
follow the diffusion of such claims to other parts of a country, a continent or the world,
understand the underlying phenomena, and to take action (if only in the form of a
warning) if needed.
If claims do already exist, this means that in fact the action will come too late. Therefore
the real challenge is to develop a watch function around emerging scientific controversies
and new technical innovations, which may potentially have great impact on the insurance
sector. This type of watch is the subject of an old debate, but appears still to be
underdeveloped in general, and particularly in the insurance sector.
From a methodological point of view the identification and the analysis of scientific
controversies and technical developments can be done in an extremely simple manner
(reading relevant journals, performing internet searches). It could easily be enriched, in
our view, with more automated analytical tools, if the data of course is available in digital
form. This will especially be necessary if great quantities of data have to be searched.
Engaging early in the R&D process
Another useful insight from innovation studies is that innovation is not only a technical
process but also very much a social one. The actors involved in innovation (users for
instance) will have an impact on the shape of its outcomes. There is no reason to think
that in the case of the insurance sector this would not be so. In selected innovations,
actors from the insurance world may want to play a more important role. It is though that
insurers can play an important role in the management of certain technological risks by
simply being there from the outset. Being involved in selected innovation processes will
14
A side remark is that new types of claims are often first observed in the United States before
crossing the Atlantic.
30
lead to a better understanding of, and a more efficient prevention of later risks associated
with the research as well as its outcomes. Insurance contracts have a strongly regulating
character and thinking about the design of possible future contracts in parallel to the
engineers who are thinking about the design of the future world will constitute a healthy
interaction and will be beneficial to the process.
For most major technological risks, especially when the innovation has not found
concrete applications or an adoption yet, however, insurance is not readily available, so
insurers cannot fulfill a role in risk sharing and reduction. Technological risks that are too
uncertain for a firm to bear may also be too uncertain for an insurance company that
typically shows ambiguity aversion [Kunreuther et al. (1995)]. Lack of actuarial data and
their public good nature, dramatic consequences that can amount to enormous costs, t he
uncertain duration of adverse effects or of the latency period, and the fact that liability
rules can change over time can deter insurers from entering the market. Moreover, the
increasing complexity of technology, which makes the causal relationship between safety
measures and risk reduction difficult to grasp, tends to exacerbate both the classical
problems of adverse selection and moral hazard.
Therefore, it could be desirable that the insurer participates – or even occupies a
proactive role – in the innovation process. This would allow the innovators to consider
the issue of risk very early in the process. Given the current involvement of insurance
companies in innovation processes for the sake of risk management, this seems a huge
task but it does not seem impossible. As an investor, insurance companies are indeed
already involved in such processes but the two roles are mostly strictly separated. Some
initiatives however, such as the one presented by GSDP within the INTEREST group, try
to associate the two approaches.
The risk of involving insurers in the innovation process is that no risk is taken any longer
by those carrying the innovation.
Engaging in the decision-making process
A third opportunity again relates to the socio-technical character of innovation and more
generally to the decision- making process around the adoption of new scientific and
technical developments. The argument is that insurers could participate more fully in
public debate around technological risk in which scientific controversy is at stake in a
similar way that other societal actors may. Following again Callon et al., the concept of
“hybrid forum” can be put forward, as opposed to traditional ways of decision- making
around technological risk. Traditionally, choices around technology have been one-off,
taken by a “legitimate” actor, and closed by either a scientific or political authority or
both. The decision model proposed by Callon et al. is radically opposed to this traditional
model. They argue that decision making in a situation of uncertainty – which is the case
that is at stake here – should be an iterative activity consisting of plural actions; it should
involve a great diversity of actors, not only “the” decision makers, or “the” scientific
community; finally there should be no definitive closure, but the process should be
reversible as well as open to the arrival of new insights and actors. In such a process –
31
which currently seems to form the exception rather than the rule – insurance companies
may well have a role alongside other actors (researchers, citizens, firms, consumers…).
Procedures which explicitly or implicitly assist in organising these new forms of decision
making, and especially different forms of public debate concerning decision making
related to science and technology have been recently pulled together in a volume edited
by Joss.
Conclusion
This paper discussed some recent insights in innovation theory. It argued that innovation
is a process in which heterogeneous actors intervene that is not following a linear pattern
from science to the market; innovation is characterised by path dependencies that can
occur rather early; finally every technical option assumes a socio-technical environment
in which it is able to function.
The second section of this paper discussed some relevant issues in technological
innovation which relate to risk and management of uncertainty. This allowed us in the
third section to propose forward some opportunities for the insurance sector to be
involved in processes of innovation. The first opportunity identified was to develop a
more systematic watchdog or early warning function in order to identify technological
developments and scientific debates which may become relevant from an insurer‟s
perspective. The second opportunity was to engage more fully and concretely in selected
innovation processes in order to understand potential and hypothetical risks more fully,
but also to make innovators more aware of potential risks and hence influence the
direction of the process at an early stage, when flexibility still exists. The third
opportunity finally is to engage more fully, not only in the “micro” R&D process but also
in more broader public debates around scientific and technological developments. In sum,
insurance companies more and more often are direct stakeholders in innovation process
since they are expected to bear some of potential losses. For this reason they should be
more interested in the organisation and the outcomes of such processes and take an active
role in the debates surrounding them.
Discussion
Innovation is a learning process, the INTEREST project should help find out about he
conditions under which insurance could play a role. The idea of more involvement of the
insurance sector in R&D would be interesting to explore, and could perhaps lead to more
certainty and removal of ignorance, on the part of both insurers and researchers. To
extend R&D in this way, risk assessments are needed, and the creation of deliberative
processes involving a wide range of stakeholders. It may be beneficial for involved
parties to learn how to cope with „interpretative flexibility‟. On the other hand,
involvement of the insurance sector in R&D processes places rather high expectations it.
The entrepreneur can be seen as a risk in itself, and there is no insurance tool to analyse
this psychological component.
32
A specific problem regarding innovative technologies is that some technologies and
associated effects are judged acceptable by some parts of society but not by others. It is
problematic to generalise about „society‟ in this context, and it might be more appropriate
to refer to societies. Also, there is always a time lag between innovative technology and
legislation, for which one needs to find a balanced solution. Because of different
legislations, it is impossible to have some global control.
There is a need to set priorities, it was mentioned that asbestos was safe until it was cut
and fragmented. However, asbestos is a special case as for a long time the potential risks
were known but nothing was done about them. Several other sources of potential risks,
such as vaccines, mobile phones and genetic manipulation, can better be cited for our
purposes.
It was suggested that the stages of innovation were broken down a bit for the purposes of
identifying were the insurance sector could contribute. Perhaps the insurance sector could
have more influence in the diffusion stage and limit negative external effects there. Areas
that need to be explored further are the governance role of the banking and insurance
systems, the issue of capital risk and insurance, and the difference between actual and
perceived levels of risk.
33
2.4 Sustainable Development, Risk and R& D Policy
Jim Skea1
Policy Studies Institute
Introduction
The three key concepts underpinning the INTEREST project are the insurability of risk,
sustainable development and technological innovation. This position paper is intended to
draw out the links between sustainable development and RTD policy, and to highlight
how the insurance sector, and insurance-based risk management mechanisms, could
either contribute to or be informed by these linkages.
The paper is a contribution to the first INTEREST workshop, which is concerned with
scoping out insurance and regulatory approaches to risk.
The paper first reviews progress in bringing sustainable development principles into
policy- making since the 1992 UN Conference on Environment and Development
(UNCED – “the Rio Summit”), highlighting key areas where further progress is needed.
The paper then moves on the assess the Sustainable Development Strategy adopted by the
EU at the Gothenburg Summit in June 2001. The following section considers the research
needs associated with the Sustainable Development Strategy, observing that the gaps
between the sustainable development and RTD policy domains remain considerable.
Finally, the paper addresses specifically where and how the insurance sector might
engage, so as to contribute to and derive benefit from the establishment of a bridge
between sustainable development and RTD policy.
Sustainable Development
The classic Brundtland definition of sustainable development endorsed at the Rio Summit
is “development that meets the needs of the present without compromising the ability of
future generations to meet their own needs”. As has been cogently argued many times,
this formulation firmly establishes the principle that the pursuit of sustainable
development must involve the simultaneous consideration of economic, social and
environmental factors. These social, economic and environmental dimensions must
reinforce each other. In addition, sustainable development must inevitably address
important questions of scale. The concept of futurity is vital and it is necessary to take a
long-term perspective. Equally, sustainable development can only be seen from a global
perspective, acknowledging however that linkages between a global view and the local
realities of citizens are vital.
1
This paper substitutes for one due to be prepared by Uno Svedin of the University of Linköping.
Due to illness, he was unfortunately unable to attend the workshop or prepare a paper. I am gratefu l
to Uno for making various working papers available to enable me to comp lete this paper and for
many insights which belong to him rather than me.
34
Significant progress has been made since the Rio summit: both EU and national policies
have increasingly come to recognise the principles of sustainable development;
businesses, through the World Business Council on Sustainable Development (WBCSD)
for example, are beginning to adopt the sustainable development concept; local
communities and local authorities have used Local Agenda 21 to pursue sustainable
development at the local scale; and major international agreements such as the Climate
Change Convention, the Kyoto Protocol and the Biodiversity Convention have been
reached, although not all parties have signed up.
Nevertheless, much remains to be done. Specifically, there has been: insufficient
integration of different policy domains; insufficient integration of different knowledge
domains; insufficient connection between the different scales at which sustainable
development operates, i.e. the global, regional and local; an imperfect articulation of the
science-policy interface; and insufficient transfer of resources between North and South.
All but the latter are relevant to the INTEREST project and are given further
consideration here.
Integration Issues
Three major issues can be discerned here. First, there is a continuing need to integrate
sustainable development into sectoral policies such as energy, transport a nd agriculture.
Second, more progress needs to be made in bringing different scientific perspectives, e.g.
bringing natural science and engineering together with social science and the humanities.
Finally, there is a fundamental need to consider systemic sustainable development issues
such as societal resilience and vulnerability. These issues were pertinent in relation to
challenges such as climate change prior to 11 September, but are even more pressing
now.
Global- local connections
Here again the needs are both scientific and political. We need to understand better how
globalisation impacts on the reality of people‟s lives at the local level. We also need to
gain an understanding of the regional dimensions of global environmental syndromes
such as climate change. For example, what are the implications of global climate change
for patterns of flooding at the regional and local levels? In responding to these challenges
we must also come to grip with the challenge of multi- layered governance. Institutions
operate at different levels – global, EU, national and local. They are connected by
complex power relationships which need to be better understood. Finally, in a globalised
world mediated by market relations we need to understand how individual lifestyles
evolve and are negotiated.
Science-Policy Relations
The critical need here is to understand differences in logic and perspective between
scientists and policymakers. They need to work together to define common challenges
and tasks and to build wider social needs into the research process. The EU is making
35
some progress in this respect by engaging policymakers in the definition and evaluation
of research programmes and proposals. This is a tangible illustration of how to meet the
need to find practical means and institutional forms to move forwards.
The EU Sustainable Development Strategy
The EU‟s Sustainable Development Strategy was adopted at the Gothenburg Council in
June 2001. It sets out a clear agenda for grappling with the remaining sustainable
development issues identified above. The key features of the programme are that it: a)
will act as a catalyst for both policymakers and public opinion; b) aims to provide clear
and stable long-term objectives providing a framework within which different actors can
operate effectively together; and c) focuses down on a small number of sustainable
development challenges which pose severe or irreversible threats to society.
To make sustainable development happen, there is a need to improve policy coherence.
Specifically, this means carefully assessing the full effects of every policy proposal
including estimates of the economic, environmental and social impacts. The groups who
will bear the burden of new measures must be identified so that adaptive measures can be
initiated. Better information is needed to underpin policy- making. However, following
the precautionary principle, lack of knowledge cannot be used to justify inaction.
It is vital to get the prices right so that both individuals and businesses have the
incentives to change behaviour and invest in technologies, products and services that will
fulfil social needs while reducing environmental pressures. One element of “getting the
price right” might be pricing risk appropriately. Here the insurance sector has a
potentially important role to play.
Investment in science and technology is essential so that sustainable development can be
pursued by reducing our use of natural resources, cutting pollution and mitigating risks to
health and safety. To the extent that new technologies are cheaper than their
predecessors, this will also contribute to the social and economic dimensions of
sustainable development. The key is to ensure that legislative and institutional factors do
not inhibit the development and diffusion of sustainable technologies. The public sector
role is to promote basic and applied research into sustainable technologies and the factors
that will lead to their adoption, and to fund benchmarking and demonstration projects.
Public procurement policies are another tool for accelerating the adoption of cleaner
technologies.
Improved communication with consumers is essential to overcome widespread
scepticism about decision- making procedures that are seen to be too remote and
scientifically driven. Policymakers need to engage in a more systematic dialogue with
stakeholders, including consumer groups, before initiating new policy proposals.
Individuals need to feel a sense of ownership about sustainable development and to feel
that their own actions can make a difference. The business sector can help by engaging
with the new corporate social responsibility agenda.
36
Finally, the global dimension, including enlargement issues, must be factored into the
policy- making process. The impact of EU policies on sustainable development outside its
own borders must be taken into account. Our actions and those of our global partners
should reinforce each other. The Rio+10 Summit in 2002 provides a unique opportunity
to take this forward.
To make sustainable development, tangible and meaningful, the EU has identified six
specific priorities which are an integral part of its sustainable development strategy:
combating climate change;
addressing threats to pubic health;
managing natural resources responsibly;
improving transport and land- use management;
combating poverty and social exclusion; and
dealing with an ageing society.
The first four priorities were agreed at the 2001 Gothenburg Council. The last two had
been agreed at previous councils.
The priorities under climate change are to develop clean energy and support relevant
R&D. Food safety, chemicals in the environment and the spread of infectious diseases are
central to the new public health agenda. Managing natural resources entails breaking the
link between economic growth, resource use and waste production, restoring natural
habitats and instituting a sustainable fishing regime. The goals under transport are to
decouple transport growth from economic activity, reduce the share of road transport and
to reduce disparities between different regions and types of community. Poverty/social
exclusion goals relate to the eradication of poverty, raising participation in the labour
market and increasing youth participation in further education and training. The ageing
society brings the challenge of pensions policies and increasing the participation of older
people in the labour market.
INTEREST case studies relate to the EU‟s priorities in terms of climate change, public
health and the management of natural resources.
The Contribution of Sustainable Development Research
Research which supports sustainable development has a number of defining
characteristics. It must be systemic in nature, i.e. it must be interdisciplinary, cut across
sectors and must take account of the way in which science is embedded in a social and
cultural context. It must also take a long-term perspective, while not losing the capacity
to guide short- and medium- term actions.
37
It must be able to operate at different scales from the global to the local. It must also be
able to identify and explain linkages between natural phenomena on the one hand and
socio-economic phenomena on the other at these different scales.
Technological systems will develop within a broader societal context. There are many
possible paths forward. Sustainable development research must be able to indicate which
forward paths will take us in unsustainable directions or pose inherent risks to society.
Technological change is path-dependent and the fact that we may become “locked- in” to
paths of development that are either sustainable or unsustainable should be recalled.
Finally, sustainable development research must be “governance-embedded”, i.e.
stakeholders should be involved upstream in the R&D agenda-setting process.
Achieving these aspirations for sustainable development research requires a suitable set
of tools for design and implementation. We need tools which:
allow us to scan imaginatively for new means of understanding sustainability and
unsustainability;
link different policy domains;
engage stakeholders, especially upstream in the R&D process;
establish conducive forms of institutional design, especially in terms of financial
mechanisms and implementation devices;
create feedback mechanisms to the policy domain, perhaps by developing
methods for synthesising research;
provide rigorous quality control; and
lead to the effective dissemination of results.
The EU is beginning to take on these challenges. The need for linkage is now recognised
at the highest political levels. For example drafts of the Framework VI RTD Programme
now contain clear references to the Sixth Environmental Action Programme. However,
there still remains some way to go. There are significant differences in perspective
between those who work in the policy domain and those work in the R&D domain.
Options which seem to researchers to satisfy reasonable scientific criteria may appear
politically unrealistic to policymakers. On the other hand, researchers may believe that
controversies in the political world as new policies emerge are evidence of a lack of
novelty or innovation. The slowly closing gaps between the research world and the policy
world are cultural as well as practical.
Engaging the Ins urance Sector
38
This synopsis of the links between sustainable development policymaking and
sustainable development research brings us finally to the question of the insurance sector
role. It is apparent that the sector can contribute in a number of ways to developing both
sustainable development policy and the sustainable development research tools listed in
the previous section. It can:
participate in fora which identify sustainable development policy needs;
participate, as a societal stakeholder, upstream in the RTD process in order to
define sustainable development research needs and priorities;
disseminate knowledge of societal/environmental risks which it already possesses,
as some companies and organisations have already done for climate-related risks;
highlight insurance-based methods and mechanisms for pooling sustainable
development risks and demonstrate how insurance-based techniques can underpin
precautionary approaches; and
re-orientate its investment practices along sustainable development principles.
The case studies that form an intrinsic part of the INTEREST project demonstrate the
specific contributions the sector might make. It can: promote the diffusion of sustainable
technologies through risk-sharing (NYMPHEA case study); more speculatively, it might
help to manage the risks associated with novel technologies and technology clusters (GM
crops); it can help to define societal vulnerabilities (flood risks); and it can promote
understanding of risk and risk transfer mechanisms.
The insurance sector has something to gain from engaging in this way. Precautionary
approaches could lead to better loss prevention as a result of better articulation between
public policy and corporate strategy. For example, a precautionary approach to spatial
planning of housing could help mitigate flood risks and reduce insurance industry losses.
The sector might also gain in terms of reputation effects from a more visible and
constructive participation in the public arena. Finally, there is also the prospect of
business development should sustainable development policies help to expand the pool of
insurable risks.
References
Commission of the European Communities, A Sustainable Europe for a Better World:
European Union Strategy for Sustainable Development, Commission‟s proposal to the
Gōteborg European Council, COM (2001)264 final, 15 May 2001
EU Council, Presidency Conclusions: Gōteborg Council 15 and 16 June 2001,
SN200/1/01 REV 1, Brussels 2001
Kates RW et al., “Sustainability Science”, Science, 292 (5517) pp 641-642, 2001
39
Svedin U, “Sustainable Development and R&D policy: The European Context”,
Presented to the conference R&D Polices in European in the Sustainability Field,
Brussels, 27-28 November 2001
Discussion
Although much remains to be done on sustainable development, the EU has made
substantial progress. Socio-economic needs have been built in since the beginning of the
Framework V RTD Programme. Socio-economic issues have been built into all
programmes, with more or less success – it has for example been central to the Quality of
Life theme. In terms of links between research and policy, DG-Environment has been
involved in the evaluation of all relevant projects.
It was remarked that it is important to focus on the balance between risks a nd
opportunities, not simply the risk side. Taking examples from the „red‟ life sciences,
people are perhaps prepared to take risks as they are more concerned with the potential
benefits from medical interventions. The example of the drug interferon in controlling
multiple sclerosis was cited. Another example of a benefit comes from the „grey‟ life
sciences where the development of „coldzymes‟ has been shown to raise resource
productivity by a factor 37,000 in certain cases.
The question was raised as to whether the INTEREST project should be focusing on the
research or development end of the innovation chain. It was thought more appropriate to
focus on the development side. By that time, hypotheses about the nature of risks can be
formulated and the key decision, where insurance has a role, is whether or not to accept
the risks.
On the investment side of the insurance business, it was observed that the industry does
tend to be very cautious in its approach and will not take significant risks or get involved
in venture capital.
In reviewing general sustainable development progress over the last ten years, it was
observed that little progress was made in the first seven years after Rio. Economic
performance was seen to be of vital importance for the concept, as it has not yet become
an irreversible norm. On North-South links, industry representatives believed that
emerging markets do not offer any potential for insurance ; the industry has done
nothing for developing countries over the last ten years. Political risk is a critical
element in considering these north-south links. It was pointed out that social exclusion
also occurrs within European societies as well as along the North-South axis. A short,
inconclusive, discussion followed as to whether poorer people fail to take out flood
insurance because they cannot afford it, or because of certain social norms associated
with expectations about the role of the State and other actors.
40
Also, there is still a lot of ignorance, with many insurance industry clients simply not
aware of what is happening with issues such as climate change. The business sector
could potentially act as a catalyst for science and policy, as scientists employed by
industry are working on the ground.
For others, this illustrated the need to get various types of actors, playing complementary
roles, involved. This was seen as being a virtue of the INTEREST project. It appeared
that the insurance sector was interested in understanding more about social risks and
values. However, on shifting risk management tasks towards the insurance sector, some
participants urged caution. Certain societal risks are by their nature fundamentally
different from traditional and insurable risks, and their management should continue to be
shaped by the public sector rather than the private sector.
Some final observations were made about the fact that the EU cannot be separated from
wider global trends, and that the EU is very much part of shaping global trends, not only
through trade and investments, but also through aid and development.
Jim Skea concluded by responding to comments about the need for focus in the project.
He underlined the fact that this is an exploratory project and the case studies serve to
exemplify wider issues. A focus on points that need to be explained and emerging
research needs is appropriate. If the team were to look at climate-related flood risk, it
would be in order to exemplify the boundaries between the private sector and the public
sector, which has a role to play through the spa tial planning system. If nuclear power
were to be looked at, it would exemplify the role of insurance in relation to a socially
contested technology. The two case studies within this workshop illustrate this. The
NYMPHEA case study shows how insurance-related mechanisms can be used to
overcome obstacles to the adoption of technologies that a priori offer sustainable
development benefits. The GM case study on the other hand concerns if and how
insurance could potentially act as a regulatory mechanism in relation to technologies
which are socially contested and whose sustainable development benefits are ambiguous.
41
3. CASE STUDIES
3.1 Insurance and Risk Management Tools to Enhance Security and Reliability
for Innovation: Nymphea Water
Jean Michel Ribouleau
Gerling Sustainable Development Project (GDSP)
Introduction
This paper concerns the role of insurance products and the insurance sector in fostering
innovative processes and their implementation in the commercial sphere. It assesses in
particular the scope for a new product referred to as a Technological Performance
Achievement Guarantee, drawing lessons from the Nymphea Water Project.
The Nymphea Water project attempts to tap water resources available from freshwater
resurgence on the sea bed. The project relates to sustainable development in different
ways, as it is thought that it will have a major positive impact on the local economy and
social context while mitigating the adverse environmental impacts associated with water
scarcity.
The project also raises a number of insurability challenges. As with all innovation, the
lack of historical data and feedback impede the traditional actuarial approach of
insurance, to the detriment of the innovation process. The possible underperformance of a
new technology entails financial consequences that could affect the business plan,
endanger third party involvement and undermine the development. In order to insure this
kind of project, the traditional actuarial approach needs to be replaced by a factual
approach, based on industrial and scientific risk assessment. New insurance products,
underwriting behaviour, and risk management tools need to be employed.
The project also considers emerging science issues, through its relationship with
members from COST 621, a European cooperation in the field of science and technical
research dedicated to groundwater management of coastal karstic aquifers.
The innovation process
From concept and prototype development through to industrial and commercial
implementation, it is widely accepted that innovation is driven by small and medium-
sized enterprises (SMEs). Innovation proceeds by anticipating and matching the (possibly
latent) expectations of potential markets, users or industry. Both the demand and supply
sides have an interest in an innovation‟s success. Starting from an idea, an invention
evolves into a patent, and the growth or creation of a company. An invention becomes an
innovation when there exists a sound potential for entering the market.
42
The scope for innovation may contrast with the weakness and inexperience of the SME,
raising issues of credibility and solvency from the perspective of the market, financiers
and investors. Various stakeholders intervene at different stages to turn the
project/product into reality. While the innovating company and the State are likely to
support and fund early stages of development, production and commercialisation phases
also need financial resources, from current assets and venture capital through to strategic
investors.
The crucial transition from prototype to commercialisation requires the build -up of
confidence in the innovation.
Due to its structural financial vulnerability, young SMEs are eager to launch innovative
products and maintain independent R&D leadership. Therefore, innovative processes
based on technological progress tend to pass directly from the research stage to the
commercial or industrial sphere. However, in applying the prototype, further
development might still be necessary, as the innovation can only be prope rly tested and
validated by the „real‟ market. Any difficulty occurring at this stage may have important
consequences for the innovative process as well as the viability of the SME.
Insurance and risk manage ment tools to enhance innovation
Risk
The generic term „risk‟ encompasses various meanings, but will be re-defined for the
purposes of this paper to refer to all financial consequences for an insured or third party
stemming from a hazardous event. The direct risk relates to physical damage to property,
whereas indirect risk relates to financial losses such as interruption risks.
From an insurer‟s point of view, a possible definition of technological risk would cover
all penalties, damages and interests incurred as additional costs by a supplier or third
party (including the client and users) as a consequence of the non-realisation of a
contract. The non-realisation might stem from failure to obtain the necessary technical
performance, or from an unpredictable technological hazard and the subsequent non-
compliance from resulting delays. Technical risk refers to direct damages and possible
consequential losses stemming from a breakdown of machinery or equipment.
Insurability
The insurability of risks can be examined from an ethical as well as a risk assessment
point of view. The former relies primarily on human considerations regarding the nature
and scope of risk, while the latter is more pragmatic and has its origins in statistical and
probability theories, making it more objective.
Insurers tend to regard innovation as a speculative procedure that will benefit the firm in
the future. Although there are no legal restrictions on underwriting innovation, it is
43
generally taken to be a “business risk”, preventing its insurability from an ethical point of
view.
If there is agreement on the part of governments and potential users that an innovation
targets sustainability, there is no impediment to insurers examining the insurability of the
associated risks. Everybody has an interest in progressive action. Societal choices and the
market will deal with the issue of moral hazard. Once this has been established, the
subsequent challenge is to ensure that the behaviour of the insured cannot affect the
frequency or severity of losses. Indeed the prerequisite for insurability is the
unpredictability of the occurrence and severity of risks, the combination of which results
in a risk event.
Obviously, the firm, even if insured, has no interest in increasing the risks it faces.
Moreover, any suspect behaviour could compromise its ability to get insurance coverage
renewed in the future, thus depriving the company of a key management tool for success.
Considering the necessary cooperation between the underwriter and the potential insured
for assessing the project, insurers (unless voluntarily for marketing reasons) are unlikely
to make an adverse selection1 . As we will later highlight in the study, underwriting is
subordinated to methodological assessment of the project and of the SME itself.
Finally, we should raise the issue of moral hazard. Indemnities should never be profitable
to the insured. Therefore the insured must retain some exposure for financing unexpected
additional costs. Excess losses (above the loss limit) and deductibles fixed in the
insurance contract should serve as efficient incentives for good behaviour and risk
management. In other words, business risk is reduced to its basic level while maintaining
an incentive effect.
Technological performance and hazard
The implementation of any innovative process is almost de facto subordinated to a
contractual supply commitment. For example, machinery for a specific industrial unit
implies essentially new processing. The transition from invention to innovation, through
supply contracts, involves obligations regarding technical performance. Whether an
innovation is insured or not, the SME needs to achieve performance. The main risk faced
by the company is non- or under-performance. The potential role for insurance depends
on the source of non/underperformance:
Incompetence (design, business management): No role for insurance.
Technical failure (materials or assembly): Use existing insurance products.
Technological hazard: New insurance products could be designed.
Of all the hazards threatening SMEs, technological hazard is at the core of the innovative
process, and could undermine all other arrangements. Our work will focus on this issue.
1
This would result in a lower quality portfolio.
44
Technological hazard (Figure 1), inherent to the innovation process, may occur at all
stages of development, however careful the studies, tests and planning undertaken in the
first conceptual and development phase. The worst outcome is the need for additional
financial resources to redo all or part of the studies, finance additional studies, and get
additional expertise or alternative equipment solutions in order to achieve successful
delivery.
Existing insurance coverage
The following section will introduce the available coverage without any a priori
assumptions about the insurance sector‟s underwriting for the innovation pro cess, which
can be considered riskier than standard underwriting. Standard guarantees intervene at the
beginning of the manufacturing stage.
Innovative and other SMEs can and must use the traditional tools on the market, which
are usually underwritten by industrial sectors supplying equipment, and fall under three
headings: installation risks (damages approach), guarantees (damages approach limited
to provider liability), and general and product liability. The core problem faced by an
innovative SME is that, apart from the most comprehensive versions of general and
product liability (consequential losses without any material damages), those insurance
products do not take into account technical performance criteria when specifying the
guarantees.
General liability covers financial consequences for third parties, including clients, from
any hazardous event, including technological hazard (as long as the consequential losses
were underwritten). Product liability encompasses all damages incurred to third parties,
clients included, from the product itself. This coverage begins after formal receipt (e.g.
delivery, service installation or ownership transfer) and implies that the technical
performance has already been verified and endorsed. Product liability insurance is
designed with respect to potential latent defects and underwrites all direct material
damages, directly related consequential loss, and insured pure consequential loss 2 from a
latent defect. Manifestation of the latent defects could involve the product breaking down
or its performance decreasing3 . Insurance liabilities never pay for repair or improvement
of the product.
Installation insurance covers machinery breakdown, whatever the origin (external or
internal), that may occur between assembly and delivery or installation of the product. It
is an extension of the traditional and well-known machinery breakdown underwritten to
cover equipment in normal service. Only material damages to the product are insured.
The insurer pays to repair the product to its original state, without any consideration of
performance and consequently does not pay for any improvements. A one-year guarantee
refers to the risk of machinery breakdown during that period, as far as the liability of the
2
Pure consequential loss may be a loss of use for the client or user, a loss of market, reputation damage.
Any prejudice susceptible to be protested.
3
The performance decrease may be total or partial, sudden or gradual.
45
supplier is involved (internal cause, incorrect user instructions). Again, the insurer pays
only to restore the product to its original state.
From hazard to risk: risk assessment procedure
The actuarial approach is incompatible with innovation, which has no history. Even
where innovation consists of a new combination of established techniques, the outcome is
nonetheless without precedent. Insurers cite lack of experience as a reason for avoiding a
pro-active stance towards innovation. Consequently, the key point is whether the
traditional actuarial risk assessment approach used by the insurance sector can be
substituted by a factual approach, based on industrial and scientific methodologies. Such
methodologies are already used to manage safety systems in sectors requiring innovative
processes, such as nuclear, aerospace and other emerging sciences.
Risk assessment, as an initial step to define underwriting and contractual parameters (loss
limit, cover granted and premium), should encompass all aspects of the project (finance,
technology and techniques, sub-contractors, management team). A trust-based
partnership between the insurer and the insured will not only optimise the client‟s risk
management behaviour, but will also underpin overall business and project management.
Partnership will also be crucial during potential claims handling or in a crisis
management period.
New tools for insurance
In this section, the scope for a possibly relevant new product, a Technological
Performance Achievement Guarantee4 , and its positive consequential impacts on
innovative project management and the appraisal of SMEs will be examined. In addition,
the section discusses how the underwriting of other traditional insurance products, using
scientific risk assessment methodology, could be applied, and whether it would be
possible for insurance to intervene at an earlier stage of the innovation process. In other
words, the insurer could intervene during the R&D phase, jointly with a state agency
concerned with innovation.
It has already been mentioned that any hazardous event may entail additional costs to
reach contrasted performance (additional studies, supply of alternative equipment), in
addition to possible contractual penalties. Insurance will cover all or part of these costs,
according to the intervention threshold and the loss limit (see Figure 2). This financial
support gives SMEs an incentive to correct any potentially hazardous situation and assure
third parties of this aspect of its solvency. The specificities of each innovative process
call for a case-by-case risk assessment and underwriting approach. However, the
Technological Performance Achievement Guarantee is a generic instrument.
4
We use the term Technological Performance Achievement Guarantee rather than Performance
Bond although the latter can be used in a generic sense to refer to an insurance guarantee tailored to
the needs of innovation. However, perfomance bond can be used to refer to other specific financial
security products and this could be a source of confusion.
46
The scope of the Technological Performance Achievement Guarantee was tested in the
French robotic sector market between 1993 and 1995, on about ten innovative processes
developed by the French company GAN. Underwriting was based solely on absolute
confidence in the reputation, know-how and reliability of the insured party. Here we
should add that a marketing approach prevailed over a more rational and scientific one,
leading to empirical project evaluation (risk assessment reached only the level usually
undertaken for well known technical risks). Nevertheless, achieved underwriting results
were excellent. In this particular case, the pilot was launched and abandoned for
marketing reasons.
The scientific risk assessment serves as a basis from which SMEs can set up relevant
integrated risk management procedures to run the project and prevent involuntary failure.
Underwriting technological performance achievement not only provides coverage but
also addresses the crucial need for the SME to demonstrate its credibility and solvency
when scrutinised by clients, banks, financiers, other insurers such as credit insurers
(financial risks) and, possibly, state innovation agencies. Even in the absence of hazard,
this guarantee is justified because of the symbol it provides for concerned parties. For
example, it could improve capital risk and current asset management, and facilitate access
to market bonds5 and corollary financial insurance guarantees (unjustified bond requests).
Nymphea Water project
Nymphea Water is a subsidiary of GEOCEAN, a specialist in offshore operations, and
was created in 2000 to exploit submarine freshwater springs. While the phenomenon of
freshwater resurgence under sea level has been known since antiquity, there have been
few attempts to tap this resource. This is an R&D project. Industrial implementation will
only take place once the exploration, contractual phase and construction have been
achieved. Each step may lead to the project‟s abandonment due to specific local
circumstances. About 42 countries, mostly in the Mediterranean, have been deemed
favourable for commercial investigation as they have potential sub-sea springs and
regularly suffer freshwater scarcity.
The Nymphea Water commercial approach is based on a self- financed exploration stage,
in exchange for obtaining the exclusive rights to explore and exploit the localised springs.
Regional projects will be designed taking into account the industrial corporate culture of
the holding as well as local constraints, partners and facilities. The project, supported
financially by the French government with three successive loans, is innovative in
concept and method. Also, throughout the research and development phase, Nymphea
gathers existing necessary competencies and technologies, not only for tapping
freshwater but also for related skills facilitating localisation, access and assessment of
spring fields.
Water resources in the Mediterranean
5
Co mpletion bond, performance bond, penalty bond, advance payment bond, retention money bond.
47
Scarce and unevenly distributed in space and time, water in the Mediterranean countries
is under growing pressure. Larger withdrawals are linked primarily to increased irrigation
and the demographic growth in the southern and eastern Mediterranean countries
(SEMC). According to Plan Bleu data6 , eight of the 12 SEMC countries now use more
than 50 per cent of their renewable water resources annually. Two of them (the
Palestinian Territories and Libya) are already using more than their renewable water
resources. By the year 2025, according to one trend scenario, ten of the 12 countries may
be consuming more than 50 per cent of their renewable water resources, with eight
consuming more than 100 per cent.
In this difficult situation, some countries abstract water from non-sustainable fossil water
aquifers or over-abstract renewable underground water possibly causing irreversible
ecosystem and aquifer degradation by more frequent saltwater intrusion. The public is
becoming more aware of scarcity issues, and in a number of areas the risk of conflict is
increasing.
A priori impact of Nymphea
Nymphea could be an economically viable alternative to conventional water resources in
coastal zones, facilitating or enabling freshwater access to areas that otherwise would not
be supplied. Currently this source of freshwater is wasted in the sea. Its salinity rarely
exceeds 5 g/l (compared to a seawater average of 30g/l), thus reducing desalinisation
costs. Nymphea should have a major positive impact on local economies and the social
context while reducing the overall environmental impact associated with water
consumption. However, it would be necessary to investigate the ecological impact of the
underwater installation and in particular of the predicted associated increase in local
salinity.
Thinking more broadly, it is expected that the project will have a positive impact on
income generation and job creation, through the development and diversification of local
activities; on improving hygiene and quality of life by supplying basic public services
such as purified drinking water; and it could perhaps even help to decrease the risk of
conflict in certain regions by reducing the pressure on water scarcity.
Insuring NYMPHEA
Proposed working plan
The first risk assessment phase focuses on the identification of appropriate risk
management options and tools. Subsequently, we will explore the potential for insurance
and transfer of the risk base, distinguishing between ins urable risks and available
coverage (benchmarking, broker) and their possible extension or improvement, and risks
that are uninsurable or insurable but not underwritten, calling for new insurance
products, both within the Gerling Group and the insurance sector in general
6
www.planbleu.org
48
Pool of competencies
We are currently gathering a pool of experts interested in participating and willing to
exchange knowledge. Their know-how and experience will contribute to the achievement
of the objectives set out above and will particularly be of valuable input with regard to
the establishment of appropriate methodologies for the risk assessment of innovative
systems. APSYS, a subsidiary of EADS, dedicated to industrial risk assessment and
safety system, have been asked to join this pool. We also expect the contribution of the
French National Polytechnic Institute of Grenoble, which has engaged in research in this
field, and the safety systems department of Nuclear EDF, for its experience and know-
how.
Methodology for risk assessment
An innovative system risk assessment (Figure 3) will be based on a risk management
approach, involving conventional methods applied to specific industrial sector, such as
aerospace and nuclear. The risk assessment is threefold:
1) Schematic determination of the different study parameters:
a. The physical process itself for capturing the fresh water
b. Stages from water pumping through to final consumption, including water
transport, storage and transformation
c. The general context of the system with environmental impact (including
population) assessment.
2) Risk evaluation in term of the risk associated with each parameter.
3) A two-stage financial evaluation. First, the expected financial cost of the
consequences, relying on the statistical probabilities of events, is appraised.
Subsequently other methods may be applied to assess the wider financial
implications.
Conclusion
The Technological Performance Achievement Guarantee is likely to increase the role of
the insurer, who could eventually become a pro-active partner in project management
rather than just a service provider. The insurer would then become an “active” actor
fostering innovative behaviour and businesses working towards success. The insurer will
not only cover the consequences of failure but will have the ability to prevent project
failure in the first place.
49
Figure 1: Representation of technological risk for the contractual phase
Figure 2: Technological hazard coverage
50
Figure 3: Risk evaluation
Discussion
Two break-out groups addressed a set of questions relating to the case study. The
following summarises their response to the questions and then captures some more
general observations made in a plenary session.
1. Is it essential to be able to form a traditional actuarial vie w of the risks associated
with the innovation? Are there othe r risk assessment approaches that can make the
risk insurable?
The main difference between GSDP and traditional insurers is that the latter employ an ex
post, traditional approach whereas GSDP uses an ex ante approach. This implies
differences in terms of skills and resources, and on the kinds of knowledge – historical
data v technical knowledge – the insurers rely on. Classical approaches are not applicable
to innovative, or emerging, technologies.
51
It is thought that the approach used for Nymphea could be transferred to other companies,
but partnerships are needed for this. It was seen as especially important to involve
reinsurers. In- house expertise is also important, if an insurer lacks expertise or
specialisation in a certain area, there will be a reluctance to insure risks from that area.
There is an important question concerning the credibility of „expert‟ knowledge and the
associated traditional approaches. For many emerging technologies, a combination of
human and technical knowledge is needed, and very many different types of damage are
important yet can be excluded from fault trees.
2. How can business and technological risks be distinguished?
First of all, there is a semantic problem as ‟technological risk‟ is an ill-defined concept.
Even so, there is a difference between on the one hand entrepreneurial risks and the risks
of technology failure, and on the other technological risks and their social co nsequences.
Risk assessment should be done at all stages, and at the entrepreneurial, business and
impact levels.
Performance, which is difficult to assess, has traditionally not been the concern of
insurers. In the Nymphea case, GSDP is interested in investing as well as insuring, so the
project may be seen to represent innovation in insurance as well as technology.
3. What type of technologies can insurance mechanis ms help to promote?
All technologies can be supported by insurance, to varying degrees. Having said that, it is
impossible to convince insurers to insure non-profitable risks. Also, the INTEREST
project is not interested in exploring insurance support for all technologies but for
particular technologies that aid sustainable development.
The relevant technologies for sustainable development relate to the environment, or to the
environmental effects of production and consumption processes. There is a strong need to
involve stakeholders for a better perception of sustainability and for highlighting the
strengths and weaknesses of different tools. Intellectual property rights are also very
important in the context of supporting sustainable technologies.
Plenary discussion
The case study highlighted the need for different assessments of the various stages; the
identification, grading and evaluation of risks all involve different techniques and
challenges. The technology performance achievement guarantees try to fill the gap left by
traditional insurance sector instruments. Risk assessment is one of the risk management
tools. Performance is considered from an unusual insurance point of view, to give
projects the opportunity to find investors.
The Nymphea Water project is the first example of Gerling approaching innovation from
the R&D stage. One concern is that it does not wish to be held responsible for
52
unpredictable long-term environmental effects. Other lessons so far from the project
relate to the difficulty of evaluating the exposure to risk, and the importance of foresight
and horizon scanning.
The capacity and willingness of the insurance industry to become more innovative itself,
in terms of expanding the set of insurable risks, will improve by a closer involvement in
innovation per se. The links between the insurance and investment functions of the
insurance sector should be further explored.
53
3.2 Regulating New Technology: The case of GM Crops
Simon Joss, Centre for the Study of Democracy, with Malcolm Eames, Policy
Studies Institute
Introduction
GM crops have frequently been at the centre of public controversy over the last five years
or so. Central to this controversy are the issues of risk and uncertainty, and how to deal
with them within the given regulatory and public policy frameworks. Consequently, GM
crops also raise a number of interesting questions relating to insurance. This case study
provides an outline of the various risk issues surrounding different GM crops, and their
relevance for insurance. The case study provides relevant questions for discussion at the
INTEREST workshop, rather than comprehensive answers and conclusions.
GM Crops – Basics
Terminology
The term GM crops is used to denote plants that grow as a result of genetic engineering,
or genetic modification (hence GM). Genetic engineering typically involves the targeted
isolation of a DNA sequence (genes or gene sequences) from the genetic material of an
organism, and the subsequent transfer of the DNA sequence into another organism. The
result is a new organism with new traits, containing DNA from two (or more) organisms.
DNA, deoxy-ribonucleic acid, is the self-replicating bio- molecule of which the genetic
information of all organisms is made. It was identified in the late nineteenth century by
Friedrich Miescher, and its three-dimensional double helix structure was discovered by
Francis Crick and James Watson in the early 1950s.
GM crops are one of the main fields of applications of biotechnology. Biotechnology
refers to a large range of industrial, agricultural and medical technologies that make use,
in one way or another, of living organisms (microbes, plants, animals) or parts of
organisms (tissue, cells, proteins, DNA), in order to create new processes and products.
Sometimes, the term is used in a way that includes conventional processes of cultivating
microbes, plants and animals (for example, beer brewing, crop breeding, animal
husbandry); sometimes, it is used more narrowly to signify the genetic modification of
organisms, or parts of organisms.
This case study focuses exclusively on GM plants, as these pose a series of new and
distinct issues for risk assessment and insurance, in comparison with traditional plant
breeding.
GM crop developments
With the development of the essential tools and techniques of genetic modification from
the 1960s onwards, plant biotechnology research has undergone significant
54
developments. While in the (early) 1980s, emphasis was still largely on basic research,
the 1990s saw a steady increase in application-oriented research and development. As far
as GM crops are concerned, there are two main areas of interest:
the cultivation and processing of food crops;
the use of non- food crops.
Around 80 per cent of research is done on food crops, such as tomatoes, potatoes, maize,
soybeans, sugar-beet and apples; the remaining part is directed towards the improvement
of crops, such as cotton, tobacco and ornamental plants. Plant biotechnology research has
also focused on genetically modified plants for use in industry (such as biodegradable
plastic, or timber production), and medicine (pharmaceuticals).
The first commercial GM plant product was „FlavorSavor‟, a tomato that was genetically
engineered to have a longer shelf life (i.e. a delayed ripening process) and taste better. It
was introduced in the US in 1994. In Europe, it was first introduced, in the form of
tomato puree, in the UK in 1995.
A recent report estimated that in 2001 around 50 million hectares (130 million acres) of
arable land in 15 countries were used for the cultivation of over 40 varieties of GM crops
(http://www.isaaa.org). This represents a thirty-fold increase since the mid 1990s. The
most widely grown GM crops are soybeans and corn (over 80% of global area), followed
by cotton, rapeseed and potatoes (for more details, see http://www.checkbiotech.org).
Processed food stuffs derived from GM crops include flour (from soya), lecithin (soya)
and fructose (corn). The US is by far the largest producer of GM crops (including
tomatoes and cantaloupes with modified ripening traits, herbicide tolerant soya, and
insect resistant cotton). Other important producers are Argentina, Australia, Canada,
China and South Africa.
Around 80% of GM crops were modified to be herbicide resistant (thus allowing
undesirable weed plants to be decimated using herbicides without affecting the GM
crops), and around 15% are insect resistant (such as Bt cotton and Bt maize).
Risk manage ment issues
Genetic modification has been a contentious issue in the scientific, political and public
spheres ever since the first experiments with recombinant DNA in bacteria took place in
the early 1970s. In the initial research phase, the scientific community itself imposed a
temporary moratorium on research (agreed at the famous Asilomar conference) over
safety concerns. In the late 1970s and throughout the 1980s, governments and scientific
organisations began to regulate research in this area. This regulatory activity has
expanded considerably since the mid 1980s with the application of recombinant DNA in
industry, agriculture and medicine.
At the centre of public debate and related regulatory activity have been the (potential)
risks of genetic modified organisms („GMOs‟), and how to manage these risks (see, for
55
example, www.genewatch.org for a biotechnology-critical viewpoint; see aforementioned
websites for a pro-biotechnology viewpoint). The following briefly discusses some of the
risks associated with GM crops, as perceived (in different ways) by various social groups.
The discussion of risks has to be seen in the wider context of the risk-benefit debate
surrounding genetic modification. It is worth mentioning that both pro- and anti-GM
interests seek to claim the moral high ground and political legitimacy offered by adopting
the discourse of sustainable development. Pro-GM groups (agricultural biotechnology
companies and their trade associations, as well as many government agencies and parts of
the scientific community) hold that the benefits of GMOs outweigh the potential risks
associated with the technology. They have argued that the development and application
of GM crops is vital to reducing world hunger and feeding future generations by for
example:
Increasing agricultural production through the introduction of high yield GM crop
varieties
Increasing the area of arable land through the introduction of salt and drought
tolerant GM crop varieties
Reducing the use of energy intensive fertilisers through developing nitrogen
fixing GM crop varieties
Reducing the use of toxic pesticides through the introduction of insect resistant
GM crop varieties
This viewpoint is contested by critics, and anti-GM groups (environmental pressure
groups, consumerists, third-world interest groups etc) have put forward a number of
specific objections to individual GM crops and the manner in which these have been
tested and marketed. They have also argued that the initial GM crops marketed by
western agricultural biotechnology companies have been designed to benefit the
developers rather than farmers or consumers.
Moreover they argue that the degree of uncertainty over the possible adverse impacts of
GM crops coupled with the irreversible nature of the technology (once released into the
environment it is self replicating and cannot be reliably contained) means that it fails the
precautionary test.
Environmental risks
A distinction ought to be made between GMOs in contained environments, and GMOs in
open environments. What is meant by GMOs in co ntained environments is the use of
GMOs in vessels that are safely sealed from the environment, such as industrial
fermenters used for the production of human insulin or human growth hormone using
GM bacteria. This use of GMOs is regulated specifically (as in the European Union
through Council Directive 90/219 of 1990 on contained use of GMOs) and is not
generally controversial in public debate any more. In contrast, the cultivation of GM
56
crops in field trials or for commercial purposes represents the use o f GMOs in an open
environment. The significance of the latter is that this use is much more difficult to
control and monitor, and the potential for large-scale risk much greater in comparison to
GMOs in contained environments. It also represents something of a catch-22 situation as
far as GM research is concerned: in order to study the risks GMOs may pose to the
environment, GMOs have to be tested in the open in field trials, but by doing so the
environment gets potentially exposed to yet unknown/ untested risks. In other words, GM
research has to rely on the environment as laboratory. The deliberate release of GMOs
into the environment has been regulated in many countries (such as in the European
Union through the Council Directive 90/220 of 1990), but the issue has nevertheless
continued to be at the centre of public controversy throughout the 1990s. As a result, a
moratorium on commercial GM crop production has effectively been in place across
Europe, in contrast to the US where public controversy has appeared much more limited.
The following lists some of the specific risks associated with the deliberate release of
GMOs.
horizontal gene transfer – contamination of wild species
Two potential risks have been debated: firstly, the unintentional breeding o f a GM
crop with a related wild plant, thus creating a new kind of weed (with the
characteristic trait of the GM plant, such as herbicide or insect resistance). Here,
the issue of safety zones around GM crop fields is in dispute (in some instances,
GM crop pollen was found several miles away from the field of cultivation).
Recent research carried out by the French national agronomic research institute
INRA showed that the relevant gene implanted in herbicide-tolerant rapeseed
could be found in wild radish, a related weed, which then became resistant to the
herbicide (Cordis Focus 2002:17). The European Environment Agency (EEA),
therefore, called rapeseed a „high risk‟ for gene flow. Secondly, the unintended
transfer of the modified genetic material from the GM crop to other species, such
as soil bacteria. Here, the possibility of a transfer of antibiotic resistance genes to
other species has been at the centre of debate (antibiotic resistance genes are used
as „marker genes‟ in the process of genetically modifying plants; see further
below).
vertical gene transfer
This refers to the process of passing on a particular genetic trait from generation
to generation. If a GM crop is fertile then the genetic modification is not limited
to the parent generation, but can be passed on to the next and following
generations (and hence, the potential risk associated with the new genetic trait is
passed on to future generations). From the perspective of the GM crop developer,
this is not desirable, as farmers could use the offspring of GM crops for further
cultivation, thus depriving the developers of revenue. Therefore, industry has
come up with so-called „terminator‟ GM crops, which are engineered so as not to
develop germ lines (pollen, seed), and which are therefore limited to one
57
generation of crops only. This would require farmers to buy the GM crops each
time they wish to cultivate them. However, due to public resistance, industry has
pledged not to use the terminator technology for the time being.
direct effects on other organisms
A GM crop contains chemical substances that distinguish it from its non-GM
parent plant (such as the proteins expressed in Bt maize that prevent the plant
from being attached by harmful insects). One possible risk of this could be the
unintended harmful effect on other, non-target organisms. For example, it was
reported that (the benign) Monarch butterflies might be at risk when coming in
contact with insect-resistant Bt corn.
insect/ herbicide resistance
Large-scale use of insect-resistant crops may lead to the target insects becoming
resistant to the insect toxin contained in the GM crops. In other words, the GM
crops could lose its resistance, as new, more virulent insects would evolve.
In the case of herbicide resistance, wild plants may develop resistance to the
herbicide developed to decimate weed (whilst being harmless to the GM crops),
thus requiring the development of new types of herbicide-tolerant GM crops.
Animal welfare
antibiotic resistance marker (ARM) genes
ARM genes are used for the process of genetically modifying plants. In order to
test whether a specific gene/ genetic sequence has been taken up in an organism,
the gene/ genetic sequence is coupled to an ARM gene, which lends the modified
organisms antibiotic resistance in case of successful modification (in other words,
the modified cell/ organism will grow on a medium containing antibiotics). It has
been argued that the consumption of GM plants might induce unwanted
antibiotics resistance in gut bacteria in cattle with harmful effects on the animals‟
welfare. Whilst the risk of an ARM gene-transfer to micro-organisms has been
judged to be low in scientific studies, it has been accepted as a possible risk.
Alternatives to ARM genes are now sought by researchers.
Human health
ARM genes
The possibility of an ARM gene transfer to gut bacteria in humans, either directly
through the consumption of GM food, or indirectly through the consumption of
animal products, has been debated. In practice, however, such transfers have not
been reported to date.
58
Allergens
The possibility of allergies induced by the consumption of GM food stuffs is
taken as a more serious threat to human health than ARM gene transfer. GM
crops may express certain proteins or other substances in different quantities than
similar non-GM crops, or they may contain new proteins/ substances that could
lead to allergic reactions in humans. A recent report by the UK‟s Royal Society
(the academy of science) called for the tightening of regulations for novel foods
(especially allergy testing), pointing out that babies are particularly vulnerable to
changes in the nutritional content of their food (www.bbc.co.uk/news - 4 Feb 02,
science & technology section).
„Social risks‟
Apart from the above types of risk, there is also what may be called the „social risk‟
associated with GM crops. What is meant by this is the perceived risk of GM crop
technology and its product at societal level, as a result of public concerns about the above
risks as well as about the socio-political repercussions of GM technology. The latter
include, for example, the impact on farming in the developed and developing world,
consumer choice issues (labelling), and ethical concerns (gene transfer across species
etc). When GM soya from the US was first shipped into Europe in 1996, this led to
considerable public controversy, not least because the industry insisted that GM soya-
flour could not be separated from non-GM flour, thus preventing consumer choice. As a
result of the controversy, retailers in many European countries withdrew GM foodstuffs,
and food products were labelled often on a voluntary basis to reassure consumers.
Also, environmental action groups have sought to destroy GM crops, arguing that they
pose a risk to the environment. In the UK, these activists have won several legal
challenges, with the courts arguing that the activist acted „in the public interest‟. The
industry reacted by scaling down its promotion of GM technology. For examp le, Unilever
decided to suspend its GM food products, and its Chairman recently stated the industry
ought to reconsider its stand on GM crops. Nestle was forced to withdraw a brand of
chocolate-bar, after it was found to contain traces of GM food.
These „social‟ risks point to the inherent cognitive, normative and practical uncertainty
associated with GM crop technology (How safe is it? Should we do it? How should we
handle it?). It is now widely recognised that the management of the social risks of GM
crops requires different means than the traditional scientific assessment of risk.
Financial Risks
In addition to the risk of direct damage to the environment or human health, recent events
have also illustrated the potential for the use of this technology to cause indirect financial
losses to third parties.
59
In recent years there have been a number of reports, particularly from the US of GM
crops performing poorly under commercial conditions. Furthermore, consumer pressure
in Europe resulted in the food industry demanding the segregation of GM and non-GM
crops. As a result segregated non-GM seed crops now often command a price premium
on world markets. This raises the possibility of GM contamination resulting in financial
damage to conventional and organic growers.
This problem has been vividly illustrated by a number of incidences of GM
contamination of non-GM seed stock sold in Europe. In one such case, in early 2000
Advanta Seeds disclosed that GM contaminated seed had been sold to farmers across the
UK, Sweden France and Germany. The contamination was claimed to have happened
when pollen from GM „Roundup‟ resistant crops was blown onto conventional oilseed
rape that was being grown for seed in Canada. Although the legal situation was
complicated, as imported oil made from this type of GM oilseed rape had been approved
for sale in the UK, there was no marketing consent for the GM oilseed rape in Europe.
Therefore the affected farmers could have potentially been breaking the law by selling
their contaminated oilseed rape crop.
In Sweden farmers were ordered to plough their crops into the ground. France was also
reported to be considering this option. In the UK 500 farms were reported to be affected,
with many major food suppliers saying that they would no t buy the crops. Friends of the
Earth estimated the value of the affected crop in the UK at approximately £1.6 million.
Advanta seeds denied liability and the UK government also refused to compensate the
affected farms (www.foe.co.uk/resource/briefings/contaminated_gm_crops.pdf).
In the UK the issue of liability has also been raised in relation to farm scale trials of GM
crops. The country‟s leading farm insurance company, NFU Mutual, has made it clear
that it will not indemnify farmers who take part in such trials in respect of liability for
any genetic pollution that may be caused.
Policy Frame work & Regulation
Regulation
GM crops, and more generally modern biotechnology, are extensively regulated.
Regulation covers research, application and commercialisation, thus includes both
processes and products. For example, to test a GM plant in the open environment, a
research organisation has to apply to the relevant public authorities. Authorisation also
has to be obtained for the commercial cultivation of GM crops. And the use of GM
products for food production is only granted following safety tests carried out by
national/international food safety authorities.
Generally, Europe (the European Union) has more stringent regulation than the US and
some Asian countries (such as China). In some countries, genetic engineering is regulated
through specific parliamentary acts (such as in Denmark), while in other countries
existing laws (relating to human health, environment etc) have been amended to take
60
account of developments in biotechnology. In the European Union, a range of Council
Directives regulate different aspects of biotechnology, such as deliberate release,
contained use and labelling (for an overview, see Durant et al., 1998).
The State has been an important driving force in the development of modern
biotechnology. Perceived as a key technology for the 21 st century, biotechnology has
been promoted heavily through specific research programmes and business start-up
initiatives. At the same time, the State has the task of regulating the technology, which
makes its role at times ambiguous, especially where there is public controversy about an
aspect of GM technology.
Increasingly, State organisations have tried to complement their scientific risk
assessments with procedures aimed at assessing the socio-technological dimensions of
biotechnology. These procedures are sometimes referred to as „risk management‟ and
„technology assessment‟. These procedures often involve a range of social actors as
discussants and assessors, including representatives of environmental interest groups,
consumer organisations, as well as „ordinary‟ citizens (for an overview of some of these
procedures, see Science & Public Policy, 1999; Klinke and Renn, 2001). They seek to
achieve a better characterisation and understanding of the perceived social benefits and
risks of modern biotechnology. GM crop technology may be deemed safe from a
scientific-technical viewpoint, but it may be considered ethically problematic. As the
social perception of risk and benefits has been shown to influence technological
innovation, technology assessment and risk management strategies aim to clarify the
socio-political dimensions of GM crop technology.
Liability
The issue of liability has long featured in the controversy that has surrounded the
commercial development and regulation of GM crops in the EU. Industry has repeatedly
lobbied against the introduction of strict liability because it feared that it would not obtain
insurance cover allowing it to bring its products to market. Likewise anti- GM groups
have lobbied in favour of strict liability for the opposite reasons.
Under the GMO Deliberate Release Directive (90/220/EC) and the Novel Food
Regulation (98/257/EC) anyone claiming to have suffered harm from products approved
under these regulations would have to seek compensation from the public authorities
rather than the companies involved. Hence a significant proportion of the risks associated
with the development of this technology have been shouldered by the public sector.
When the Deliberate Release Directive was recently being revised there was a concerted
although ultimately unsuccessful attempt by members of the European Parliament to
introduce a strict liability clause making biotechnology companies liable for any harm to
human health or the environment. At the time the Commission argued that this issue
should be dealt with under its forthcoming „horizontal‟ proposals for an environmental
liability regime.
61
The Commission formally adopted a proposal for a Directive on Environmental Liability
with Regard to the Prevention and Restoration of Environmental Damage on the 23
January 2002
(http://europa.eu.int/rapid/start/cgi/guesten.ksh?p_action.gettxt=gt&doc=IP/02/127|0|RA
PID&lg=EN ). These proposals will now provide the focus for the ongoing European
debate over the division of public and private liability for the environmental risks
associated with GM crops.
Issues for insurance
In Europe, current regulation of the deliberate release of GM crops (either for research/
testing, or commercially) does not include civil liability or insurance for environmental
(or social) harm. So, any potential harm, even large-scale, would presumably have to be
dealt with by the State, and more widely, the general public.
For insurance, the following problems/challenges exist:
horizontal gene transfer: unintended transfer from GM crop to other organisms
(problem for product liability)
vertical gene transfer: the genetic trait is passed on from generation to generation
– that is, the „product‟ is not limited (the „terminator‟ technology would be an
exception, but industry had pledged not to use it commercially)
hazards: unknown/unquantified potential environmental, health and social hazards
involved in GM crop technology (problem of quantification of environmental
damage)
difficulty of proving damage given the complexity of the technology and its
interaction with the environment
possibility of multiple source of damage (for example, if a GM tomato is found to
be harmful, is it the developer, the farmer, the retailer or the restaurant that is
liable?)
if R = S x P (whereby S is scale, and P is probability of harmful event) is used as
the basis for assessing whether something is insurable, then GM crops pose a
problem, as both S and P are unknown quantities. Even if P may be very small, S
could be potentially very large (in an open system)
Conclusions
This case study illustrates that biotechnology is a contested innovative new technology,
that potentially offers not only significant benefits but also significant risks, both of
which would have important sustainability implications
62
One may argue the absence of a strict liability regime has been an important factor in
facilitating the introduction of GM crops into the European market. Furthermore, the
protection offered to developers by the GMO Deliberate Release Directive and the Novel
Food Regulation has essentially shifted much of the potential risk from any unexpected
hazards arising from these products onto the public sector. Nevertheless novel liability
problems associated with the GM contamination of non-GM crops have recently begun to
emerge.
In the absence of a counter- factual – a case study of a market that has operated strict
liability – it is impossible to prove how much difference this has made. However, there
are indications to suggest that the insurance industry would have been unwilling to
underwrite the currently unquantifiable risks associated with introduction of this
technology if strict liability had been required.
Questions
are the methods of risk management and technology assessment developed in the
last decade or so by State organisations potentially useful for the insurance sector,
as a complement to traditional (scientific) risk assessment?
could the State adopt assessment strategies from the insurance sector to manage
GM crops-related risks?
are insurance-related approaches not applicable because of the strategic nature of
biotechnology research and application, and the nature of risks involved, and thus
there is no (or only a limited) role for insurance in this technological field?
could the involvement of the insurance industry in biotechnology R & D help
innovation?
References
Durant, J., Bauer, M. W., and Gaskell, G. (Eds.). 1998. Biotechnology in the Public
Sphere. A European Sourcebook. London: Science Museum.
Cordis Focus, 2002. “European NGOs react to new research on GM crops”. Cordis
Focus, nr.194 (8 April 2002): p 17.
Klinke, A., and Renn, O. “Integrative Risikopolitik. Ein Konzept zur Bewertung,
Klassifikation und Management von technologischen, gesundheitlichen und
natuerlichen Risiken.” TA-Informationen 4 (2001). Stuttgart: Akademie fuer
Technikfolgenabschaetzung in Baden-Wuerttemberg: pp 4-13.
63
Science and Public Policy, 1999. Special issue on public participation in science and
technology. Science and Public Policy, 26 (5): 289-380.
http://europa.eu.int/rapid/start/cgi/guesten.ksh?p_action.gettxt=gt&doc=IP/02/127|0|R
APID&lg=EN
http://www.checkbiotech.org
http://www.bbc.co.uk/news
http://www.foe.co.uk/resource/briefings/contaminated_gm_crops.pdf
http://www.genewatch.org
http://www.isaaa.org
Discussion
Two break-out groups addressed a set of questions relating to the case study. The
following summarises their response to the questions and then captures some more
general observations made in a plenary session.
1. Would an insurance/liability-based approach to regulating GM crops have been
possible? What would have been the implications for rates and directions of
innovation?
The key question is on whom the liability would fall – on manufacturers, distributors,
farmers, retailers, or all at the same time?
The role of liability under law was discussed. What would be needed for a court action to
succeed against a grower of GM crops (or related actor)? Possible solutions suggested
included compensation funds (which might mean admitting liability) or insurance for
other producers, analogous to business interruption insurance. Reference was made to
compensation to farmers for foot and moth disease (FMD) where there had been a
reaction to compensating producers from public funds, especially if farmers might have
been able to anticipate or manage some of the risks.
A problem here was seen to be that underwriters have a limited understanding of
technology. A suggestion was made that insurance could take a quasi-regulatory role in
relation to GM – for example if insurance was made compulsory, insurers could
64
effectively set rules on such things as crop separation. If these rules were breached, it
need not pay out to affected third parties. Different insurance levels might be set in
different regions to reflect underlying differences in risk. However, it was pointed out
that an insurance/liability based approach does not constitute regulation in itself.
The basic choice appeared to be between a compensation fund and liability insurance.
The Bhopal example was cited where the Government of India sued Union Carbide and
then distributed the compensation to affected parties.
There is the possibility of “no-fault” compensation schemes for third parties, raising the
issue of whether successful biotech companies should or would pay. Insurance companies
were unlikely to be interested.
The question was raised as to where the onus of insurance should be placed. Should
organic farmers insure themselves against GM contamination, or should GM producers or
farmers insure against this?
The question was raised: if public sector regulations were breached, could an affected
still take civil action against a GM producer? There is an increasing tendency to take civil
action on top of criminal proceedings where the burden of proof is different.
The insurance sector felt the GM-related liabilities were too dangerous to take on, fearing
especially that government might change the rules after the risk had been priced, and so
the limitations to the cover would remain unclear.
It was pointed out that the question concerned the finer details of an issue where the
larger picture remains uncertain and undecided. It was felt that policy on GM crops must
be a societal decision, and to reach this sort of decision for a new technology whose
application has elements of irreversibility, a societal technology assessment could be
helpful. Insurance is just a tool, and should not be in charge of setting the agenda.
As far as rates and direction of innovation were concerned, it was thought that the
availability of insurance/liability-based approaches would have made little difference to
the technology development and application, as the controversies associated with the
technology have not been around the financial losses but around the other risks.
2. Are the methods of risk manage ment and technology assessment developed in the
last decade or so in the public policy domain potentially useful for the insurance
sector as a complement to traditional risk assessment?
It was pointed out that in the area of GM crops, actual performance is of vital importance,
but some technology assessments do not take this into account. Originally, technology
assessments were largely concerned with performance, employing economists alongside
engineers and scientists, the social assessment of technology is a more recent, and only
partial, phenomenon. If the actual performance is zero, then it does not matter how good
the technology is. NGOs and citizens have been raising the issue of performance instead,
when the official health and safety assessments have not addressed them.
65
It was pointed out that quantification is a precondition of insurance, and these methods of
risk management and technology assessment do not help directly with that, but perhaps
they could help communicating clearly that insurance cannot change effects, only politics
can.
The discussion turned to how these issues relate to the broader political context. The view
was expressed that in the UK there is no effective political opposition because the bi-
party system has effectively collapsed, and this has perhaps weakened the democratic
debate. There was disagreement on this issue, an opposite view offered was that the bi-
party system has not collapsed, and the reason why there is no effective political
opposition is because all major UK parties are by and large supportive of biotechnology,
with GM opposition not organised in a party-political sense. A party in power with a
secure majority may feel less need to engage with citizens. But then on the other hand,
there are emerging and alternative forms of debate, eg GM crop activists, protest
movements etc. The issue has also become very emotional, which in itself is not a bad
thing but has tended to oversimplify the debate, and in the UK at least, there is a clear
polarity between NGOs/activists on the one hand, and corporate interests on the other.
There are two functions of participatory risk assessment methods which must be
distinguished. First, the methods can be used as analytical, or qualitative, tools, to gain an
understanding of social issues. The second function is as an aid in political decision-
making. The insurance industry could employ these methods to gain qualitative
information on uncertainty and risk perception. While most insurers will claim they have
no social and ethical concerns, the risk assessment, from the view of the insurer, should
still include some understanding of other (human) dimensions to the risks debated.
However, there might be a role for insurance as participants in these assessments, where
perhaps they could contribute with knowledge and insights. By participating, insurers
would also get involved at an early stage.
3. How applicable are insurance-based approaches given the strategic nature of
biotechnology research and application and the nature of the risks involved?
It was felt that insurance-based approached were not particularly applicable in this area,
because the risks and benefits involved are of a wide societal nature. Biotechnology is not
an insurance problem but a societal problem. Because of the issues and characteristics of
irreversibility, uncertainty and equity considerations, insurance should not be a substitute
for socio-political processes.
Insurance can never be the whole solution to anything, but could possible become part of
the solution. In New Zealand, the Earthquake and War Risks Commission, concluded that
the state should insure up to a certain point, and the insurance industry beyond this point.
Insurers agreed to this because they could control their exposure, ie how many houses
they underwrote. With new risks, the problem is that insurers do not know when to stop
underwriting.
66
This led to a discussion of cultural issues; the Dutch non-availability of flood insurance,
the differing vehicle insurance agreements in different countries, and the fact that in
Kobe, only 3% of losses were covered by insurance, as Japanese earthquake insurance is
seen as an insult to the gods. Instead there has been in Japan a focus on prevention,
through building codes. The examples highlight the need for prevention as well as
insurance; insurance does not prevent risks.
Another problem insurance is not addressing is the public‟s perception of risk, and its
diminishing willingness to accept losses. This culture and change in public perception,
will affect how insurers will deal with a lot of these issues. In this context, it was pointed
out that while rights tend to be well-known, duties are not. Similarly, we want profits but
not their associated risks. Today‟s society has become very specialised and the two have
been (artificially) separated. This lack of connections must be dealt with.
4. Could public policy adopt assessment strategies from the ins urance sector to
manage risks associated with GM crops?
This question was felt to be unanswerable because insurance companies cannot take on
these risks. However, the hypothesis was made that a liability existed. Under what
conditions would insurers enter the market? One pre-condition was felt to be that
government would ultimately underwrite the risks. The example of Pool Re, the company
established to insure terrorism risks in the UK, was cited. Here, the government
guaranteed the capital of that company from the Exchequer contingency fund. However,
this would eventually have to be paid back by insurers in the event of a large pay-out. It
was noted that all precedents such as Pool Re were voluntary in nature.
A compensation fund could be established through a levy on companies (e.g. GM
producers). In this case there need be no insurance company involvement.
It was noted that terrorism and GM were not necessarily comparable cases. The definition
of the insured event was clear in the case of terrorism, whereas cause-effect chains and
the nature of the insured event were highly uncertain in the case of GM.
Plenary Discussion
It was established that whether one is for or against GM crops, they are already in the
environment and that risk must be managed. There are two major problems in evaluating
this risk. The first is that we have a very limited understanding of complex ecosystems,
so large-scale unforeseen ecosystem effects are possible. Second, there might be long-
term health effects on the human population some generations down the line.
The issue of economic risk or loss that might be suffered by non-GM farmers (eg organic
farmers, bee-keepers or conventional farmers who are under pressure to remain GM free)
was felt to be important, and represent a gap in legislation, with no case law to build on,
and policymakers‟ thinking only at the early stages. Could the insurance industry help
resolve this gap?
67
It was asked if the insurance industry would voluntarily take on any risks associated with
GM, but mainly there was scepticism that the insurance sector could in fact take any role
at all, primarily on account of their inability to calculate a premium. Insurance is based
on one-year contracts, which do not lend themselves to the cover of long-term risks.
Where it has been difficult to get risk cover from the insurance sector, some industrial
companies have set up their own pooled insurance arrangements – for example in relation
to asbestos.
It was pointed out that there is a danger of fabricating policies that do not mean anything.
Many insurers have learnt lessons from asbestos, PCBs, and other „miracle‟ products,
about diving in to a new market before gaining adequate knowledge about the effects of
the products or technologies involved.
Others cautioned about drawing analogies between GM and the risks associated with
asbestos and nuclear power. In the asbestos case, the problem had been the ignorance of
users about potential risks. In the nuclear case, the risks discussed were those associated
with a single catastrophic accident. In the GM case, the risks fall into a number of
categories – health, environmental and economic – and there is also a lot of public
concern and fear. The advice received by the UK government so far concerns only the
risks to human health, with the experts‟ consensus being that there are no long-term
health effects from GM crops.
It was pointed out that the health risks associated with BSE, for example, had not been
anticipated. Two key issues here are whether experts are confident in their judgments
about health risks being negligible, and whether there is wider public confidence in these
judgements. Government advice was felt to be simplistic in its clarity, and the similarities
between BSE and thalidomide and now GM crops and MMR vaccinations were noted.
Another issue is that there is a substantial amount of public opposition to GM crops,
which highlights the need to incorporate the viewpoint of potential users into the
discussions. Many stakeholder discussions ignore these even though a reluctance on the
part of the potential users to embrace a new technology means there would be no
business case for such a technology, and these concerns clearly influence the economic
risk of the technology.
It was stressed again that new technologies bring benefits as well as risks, which is
especially apparent in the „red‟ life sciences where there are clear tangible benefits for
individual citizens, although the regulation (EU Directive 2001/18) does not specifically
mention benefits. It was thought that perhaps there are fundamental trade-offs between
the pace of innovation and the degree of acceptance of the societal risks, whether health-
related, environmental or economic.
Another view was that insurance is concerned with the voluntary transfer of risk from one
party to another. Public policymakers could take a larger view of the common good and
transfer risks from one party to another through legislative or regulatory devices. Does
68
allocating more risk management to the private sector, with its need for quantification
and premium calculation, shift the balance toward the precautionary end of the spectrum,
and slow down or even stop the pace of innovation in some areas? Through legislation,
policymakers can create space for innovation to take place by limiting liability and
dispersing some of the risks of new technologies more widely in society. It could be
argued that such a transfer of risks by the state would be appropriate, and legitimate, as
the state would be acting for the „common good‟.
The insurance sector is however concerned about two things. First, the expert advice
heeded by governments can turn out to be wrong, as in the case of BSE and thalidomide.
When this happens, insurers, as well as the state, pick up a large share of t he costs.
Second, policymakers often change the rules halfway through the game, creating
regulatory risks which hinder insurance companies from entering the market, thus
compromising the profitability and perhaps even viability of the whole sector.
Another industry representative raised the fundamental question of equity. This was seen
to be essentially an ethical issue. Groups of companies for example might have joint
liability. Individual organisations might have to pay for risks they have not directly
imposed on society. It was generally assumed that insurance companies are “eternal”
organisations. In relation to long-term risks, how can one generation of investors transfer
risk to another where there is a lack of knowledge about the nature of the risks? This
touches on fundamental interpretations of the sustainable development principle.
The regulatory systems developed for GM crops, pesticides, hazardous chemicals etc,
have very specific tests and cover only very specific risks. A final set of points related to
the need for horizon-scanning activities by insurers, to identify new risks and the use of
deliberative methods to engage stakeholder groups with wider perspectives. This was
especially important where risks were unquantifiable. However, it was cautioned that
there was an obligation to clarify whether deliberative processes were used to fulfil a
scientific purpose or a political purpose. These should certainly not be confused in any
individual process, and the INTEREST project should focus on the scientific aspects.
Finally, it was pointed out that insurance concerns the safety or risk of something,
whereas sustainable development issues need to consider safety as well as usefulness.
There is a need for balance between the two, and real societal choice.
69
4. CONCLUSIONS
Final discussion
The Nymphea Water project case study highlighted the need for different assessments of
the various stages; the identification, grading and evaluation of risks all involve different
techniques and challenges. The technology performance achievement guarantees try to
fill the gap left by traditional insurance sector instruments. Other lessons so far from the
project relate to the difficulty of evaluating the exposure to risk, and the importance of
foresight and horizon scanning.
The capacity and willingness of the insurance industry to become more innovative itself,
in terms of expanding the set of insurable risks, will improve by a closer involvement in
innovation per se. The links between the insurance and investment functions of the
insurance sector should be further explored.
In the case of GM crops, the general conclusion was scepticism that the insurance
industry would be interested, or useful, in getting involved at an early stage or with the
regulatory aspects of this technology.
There are two questions to consider, first, if there is a role for insurance in this area, and
second, if the insurance sector would be willing to play that role.
The role of insurance is not to tell society what societal strategies are or should be, but to
insure risk. It deals best with quantitative and priced information, and cannot for example
deal with species extinction, as no price has been put, or can be put, on this. Insurers will
only get involved if the risks can be quantified, if there is a limit to liability, and if that
limit is likely to remain stable. In the past, insurers have felt they could not take on, for
example, the risks of burglary (100-120 years ago), or subsidence and flood. This has
changed as knowledge has become better, and quantification had become possible, or, as
in the case of subsidence, government intervention has made insurance mandatory. The
government can no longer dictate terms to the insurance sector, as the insurance market
has become too uncertain.
Key points arising from the workshop
The INTEREST case studies presented at this workshop have illustrated two possible
roles for insurance in relation to innovation and technological development with
sustainability implications.
1) where there is a prima facie case that a new technology will contribute to
sustainable development, and there are barriers to its uptake in the marketplace,
insurance mechanisms can facilitate its adoption by sharing technological risks
and providing a more secure financial framework for developers. However,
insurance cannot address the entrepreneurial risks associated with innovation. The
NYMPHEA case study exemplifies this situation and offers opportunities for
replication.
70
2) Where the sustainability or otherwise of new technology is contested, the role
explicitly or implicitly assigned to insurance by the State, through regulations and
the use of liability regimes, and the consequent implications for the allocation of
risks to different actors, will determine the degree to which insurance acts as a
mechanism regulating the rate and direction of technological change. The GM
crop case study illustrates this situation.
It is clear that the role of insurance in any particular circumstance is dictated very much
by the legal framework in which the sector operates. The State may mandate insurance
(car insurance), forbid it (flood risks in the Netherlands) or establish liability frameworks
which either limit the role of insurance or give it a key role in determining whether risks
are accepted at all.
The GM case study also illustrated that insurance has a role to play in relation to
financial risks and risks borne by people, where a monetary value can be placed on
those risks. (e.g. health risks). Insurance cannot deal with pure environmental risks where
individuals or groups are not directly affected and where there are no property rights
associated with the asset at risk.
It has been shown that the capacity to accumulate actuarial evidence is not strictly
necessary for a risk to be insurable. Quantification of risks is needed, but this could be
based, as the NYMPHEA case study shows, on an ex ante analysis of risk exposure and
the probabilities of insurable events occurring. What this does require is a much closer
relationship between insurers and insured parties with the insurer acquiring and
developing specific technological knowledge of the activities which the insured
undertakes. It has been noted that insurance companies are not traditionally repositories
of technical expertise. Expanding the universe of insurable risk in this way would require
an expansion of insurance sector competences.
Boundary of state/private sector. Insurance mechanisms involve the voluntary transfer of
risk from one party to another. Both parties, the insurer and the insured, will in principle
feel themselves to be better off as a result of the risk transfer. On the other hand, when
the State transfers risk from one party to another through legislation, or the
authorisation/licensing of certain activities without a liability regime, some parties will
involuntarily bear certain risks while others benefit. This may help to promote what is
seen to be the common good. The State is legitimately entitled to take such actions in a
way that private companies are not. Where new technologies expose some parties to risk,
while offering wider and more distributed benefits, the State may decide that the
“positive externalities of innovation” outweigh the potential costs to affected groups.
There has been a strong presumption in all discussions that the role of insurance in
relation to major “societal risks” is limited and the State will always have a primary role
in regulating such risks. There was a view that the insurance sector would not regard such
risks as insurable and would decline any opportunities to become involved. This emerged
71
primarily from the GM crop case study. This judgment was arrived at by “gut feeling”
rather than detailed analysis.
The situation where the State offers no specific controls over a new technology other
than to define strict liability to third parties provides an interesting if hypothetical
counterfactual to the reality of regulation. Technological development would then take
place only in such a way that affected parties could be compensated for any losses that
they suffer and the risk were deemed to be “insurable” by the insurance sector. This
would depend on the capacity of the sector to quantify the risks involved (whether by
actuarial means or ex-ante estimates of risk exposure) and whether the risks were limited
temporally and spatially. Some would see this hypothetical regime as representing a true
application of the “precautionary principle”. The way it would take account of the
distribution of risk across society might also be seen as implementing the equity aspects
of the sustainable development principle. In general, major technological advances with
wide societal implications (such as GM) would almost inevitably be severely inhibited by
this type of regime.
Using a strict liability regime as a hypothetical benchmark, highlights the fact that the net
effect of most regulation which limits liability, notwithstanding safeguards, is to
underpin innovation, sustainable or otherwise, by allowing developers to appropriate the
benefits of innovation while transferring risk to other parties.
Although there was some discussion about the insurance sector becoming involved as a
stakeholder “upstream” in the innovation process, the industry shows little interest in
“blue skies” research. It can be concluded that the role of insurance is “downstream” in
the innovation process and is particularly relevant to the demonstration, adoption and
diffusion of new technologies.
There is little practical connection between the insurance sector’s investment activities
and its underwriting activities. In principle, risk management via underwriting could
affect the “riskiness” and returns on investment (e.g. the underwriting of flood risks
could affect property investments) but in general the link is likely to remain tenuous. In
any event, the investment strategies of insurance companies tend to be conservative (e.g.
they do not engage in venture capital) and avoid risk.
72
Annex 1: List of Participants
Christophe Barnier Apsys-Eads
Kristina Dahlstrom Policy Studies Institute
Graham Davis DEFRA
Bas de Laat Technopolis
Gerry Dickinson City University Business School
Andrew Dlugolecki Consultant
Marion Dreyer Center for Technology Assessment, Baden-Wurttemberg
Malcolm Eames Policy Studies Institute
Soraya Fahmy Technopolis
Elie Faroult CEC-STRATA
Tim Hoad DTI Future & Innovation Unit
Simon Joss Centre for the Study of Democracy
Anna Kingsmill- Vellacott BRE
Dirk Kohler Gerling
Norman Moore QinetiQ
Sophie Pardo Apsys-Eads
Jean Michel Ribouleau GSDP
Peter Simmons Centre for Environmental Risk, UEA
Jim Skea Policy Studies Institute
Walter Stahel Geneva Association
73
Annex 2: Workshop Agenda
Thursday 28 February 2002
11.00 Welcome and Introduction
11:15 The INTEREST project Jim Skea, PSI
11:30 Insurance, Risk Management and Culture Walter Stahel, Geneva
Association
12:15 Innovation and Insurance Bas de Laat, Technopolis
13:00 Lunch
14:15 Case study 1: Insurance and Risk Management Tools Jean-Michel Ribouleau, Gerling
to Enhance Security and Reliability for Innovation
15.00 Break-out groups
16:00 Tea
16:30 Report-back and discussion
- lessons from the case study
17:15 Sum-up and introduce the following day
20:00 Dinner
Friday 1 March 2002
09:00 Sustainable development, risk and R&D policy Jim Skea, PSI
09:45 Review: when is a risk insurable? Gerry Dickinson, City
University Business School
10:30 Case study 2: GM crops Simon Joss, University of
Westminster
11:15 Coffee
11:30 Break-out groups
12:45 Lunch
14:00 Report back and discussion
- lessons from the case study
14:45 Led discussion – what did the case studies tell us PSI
about insurability? Innovation potential?
Contributions to sustainable development? Cross-
cutting issues?
15:30 Sum-up and next steps
Tea
74
Related docs
