1 The Geneva Papers on Risk and Insurance Vol. 28 No. 3 (July 2003) The Role of Insurability and Insurance by Walter R. Stahel What is the past and present role of insurance in relation to innovation which promotes, or challenges, sustainable development? This is one of the key questions studied in the INTEREST project, which is presented in one of the contributions to this edition of The Geneva Papers. 1 The concept of the insurability of risks can be defined as the natural borderline between the market economy and nation states: risks that can be insured need not be legislated; uninsurable risks, however, have to be dealt with by nation states. But insurability can also be regarded as the link between technological innovation and sustainable development, as shown in Figure 1: Insurers Technological innovation Concept of insurability of risks Policymakers Sustainable development Figure 1: The insurability of risks at the centre of a balanced societal development A number of examples can illustrate the range of circumstances in which different actors have influenced technological innovation in the past. They fall into three groups: insurance has influenced technological innovation, both by promoting it and by steering it away from certain directions; nation states have played a similar role, by promoting or forbidding technological innovation; Vice Secretary General, The Geneva Association. 1 Insurability frontier between insurers and policy makers in choosing sustainable technologies? The examples described in the present article are based on newspaper reports and actual events which have occurred in recent insurance history and hence are largely unfamiliar to experts outside the field of insurance. Published by Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK.
2 THE ROLE OF INSURABILITY AND INSURANCE 375 economic actors have taken measures to prevent or reduce risks for economic reasons or as good corporate citizens. One of the common principles of these examples, the objective of loss minimization, is also the link with sustainable development: loss prevention is a prevention of social hardship, of wasted resources and of economic losses. Promoting technological innovation, which enables economic progress without losses in social and natural capital, seems an appropriate objective for our times, and insurance has a key role to play in it. In some cases, however, the most efficient strategy may be non-insurance. 1. Loss prevention through non-insurance: the self-responsibility of actors 1.1 The political decision to forbid flood insurance in The Netherlands A catastrophic flood on 1 February 1953 was caused by a major break in one of the dams; the water covered about one-third of the Dutch national territory. The loss of life and the damages caused by the flood were tremendous: hundreds of people as well as thousands of heads of cattle drowned; and the loss of property was truly catastrophic. As in most disasters, insurance payments covered only part of the losses. As a consequence, the Dutch parliament passed a law that prohibits flood insurance. The objective was to make it clear that only engineers can prevent catastrophic floods, not insurance, and that regular maintenance of and adaptations to the dams are the best strategy to prevent a similar disaster from occurring again in the future. In contrast to mandatory national flood insurance, this approach strongly promotes preventive actions. The definition of clear responsibilities reduced the danger of a second major flood and led to the development and application of new technologies and maintenance strategies. 1.2 Excluding tailgate accidents from insurance coverage in California In the early 1990s, the acceptance of California s proposition 103 reduced annual car insurance premiums by 25 per cent. Although most insurance companies threatened to withdraw from the Californian market if the proposition was accepted, in practice only one insurance company did so. The others decided to reduce the claims paid by a similar proportion through the introduction of two measures: The use of second-hand spare parts to repair damages in insured accidents, on the condition that the parts were younger than the damaged car. This led to a number of cases being brought by automobile manufacturers against insurance companies, which could prove through tests that the safety of these repairs was not impaired by the use of secondhand parts. The cases were settled out of court. The exclusion of rear-end accidents even under fully comprehensive policies, on the grounds that such accidents can be prevented by the driver by keeping a sufficient distance from the car in front. The first strategy is highly relevant to sustainability, as it promotes a market for used parts in a regional economy. The second strategy should lead to accident prevention instead of claims handling by insurance. After an initial success period, police seem to have stopped reporting rear-end collisions as the cause of accidents, due to social concerns about the large-scale financial implications
3 376 STAHEL for people involved in such accidents. Loss prevention strategies that include the attribution of personal liability for damages as a punishment for faulty human behaviour may therefore not be executed for social reasons. As a consequence, the incentives and potential impact of these strategies may be lost. 1.3 Reduced insurance for high-speed car accidents The German government has stated a recommended speed limit of 120 km/h on motorways. Implicitly, the degree of insurance cover is related to this recommended speed, decreasing to zero at a speed of 200 km/h. A typical case involved a car driving at 200 km/h in the left lane. When it approached another car driving at only 120 km/h, the driver had to brake hard and lost control of his vehicle. The resulting accident wrecked the fast car. Despite the fact that it was covered by a fully comprehensive insurance policy, the insurance company refused to pay for the loss, arguing that the accident would not have happened had the driver respected the recommended speed limit of 120 km/h. Refusals by insurance companies to pay for damages caused by high-speed cars have led to several appeals to the Federal Court in Karlsruhe. But the latter has upheld the right of insurance companies to reduce or cancel their cover on the grounds that by driving faster than the recommended speed limit, the driver takes a voluntary risk. Neither the insurance sector, nor the government, nor drivers associations give these issues much publicity, as they do not see their role in providing unwanted education for their clients. These facts are therefore little known in Germany, and the loss prevention potential of this strategy is lost. 2. Insurance influencing technological innovation paths 2.1 Limiting the size of ultra large crude carriers In the 1980s, the trend towards ultra large crude carriers ( ULCCs ) was driven by economies of scale, the main factor being crew costs. The construction of a new generation of ships of 1 million dwt, 2 up from 500,000 dwt, was cancelled after ship owners realized that the incremental gains achieved were smaller than the increases in insurance premiums. The calculation of these premiums was heavily influenced by exposure to environmental impairment liability, with dis-economies of risk in case of shipwreck. Market demand for the 500,000 dwt ships later declined, as they were too large to navigate on some of the major short cuts, such as the Suez Canal, or to unload in many of the major ports. It can thus be said that the fairly priced insurance premiums for ULCCs prevented shipowners from investing in a non-viable technology. 2.2 Influencing underground hydrocarbon tanks In the 1970s, public authorities became concerned about soil and groundwater contamination through spills from underground tanks, generally single hull constructions, 2 dwt ¼ dead weight tons
4 THE ROLE OF INSURABILITY AND INSURANCE 377 to store liquid hydrocarbon products. The solution chosen was mandatory pollution insurance for underground tanks. For years, there was no actuarial evidence about leakage. Insurance premiums fell each year in the absence of any claims. However, when the first tanks were removed, it was found that they had almost invariably developed tiny fissures through which hydrocarbon had escaped undetected over long periods of time. The clean-up costs were covered by the insurance contracts, but as a result of the claims paid insurance premiums shot up. Another result was new legislation allowing only underground tanks equipped with double skin and vacuum control of the void an expensive solution. Single hull tanks situated above ground replaced many underground tanks. This second solution has turned out to be of a high risk in major floods, often leading to nat-tech disasters (natural disasters exacerbated by technical factors). This case shows that actuarial evidence is not strictly necessary for insurance. There was no actuarial data on subsidence and leaks from petrochemical tanks when insurance was made compulsory. Insurance can thus help to introduce safer technologies at a definite cost for the client and without financial risks to the state. And this case also shows that new solutions may entail new risks, too. 2.3 Promoting innovative products and processes in life sciences Both users and suppliers of innovative products and processes accept risks if the benefits are visible and insurance cover is available. This is exemplified by the use of the drug interferon to combat multiple sclerosis, and the development of enzymes for use in process technology, such as coldzymes for washing. The sustainability impact of enzymes substituting for chemical catalysts can be incredible, enabling increases in resource productivity by factors of up to 37,000. Insurance has covered these new products of the red and grey life sciences without any need for state guarantees or intervention. In the case of green life sciences, however, European states have preferred to legislate rather than to let insurability decide which technologies are insurable and thus acceptable to the market. 3. States mandate insurance to protect society 3.1 Third-party car insurance Third-party liability insurance for automobiles is mandatory in almost all countries of the world. This helps to protect innocent parties, but it is also needed because people tend to overestimate their own driving ability. In one survey, no fewer than 75 per cent of car drivers estimated that their driving capabilities were above average. From an insurability point of view, car insurance is an almost ideal risk, as each year there is a large but random number of accidents with a limited liability in each case. In many countries a bonus malus system operates, benefiting drivers with few accidents and punishing bad drivers. The system can be linked to a person (as in the U.K.) or a car (as in continental Europe). The rating can be based on car models (as in Germany) or engine size, or drivers characteristics, or a mix of these. In conclusion, states can use, and have used, insurance as an efficient tool to protect vulnerable third parties and create disincentives for faulty actors.
5 378 STAHEL 3.2 Compulsory insurance of pressure vessels After the first major accidents with pressure vessels in the age of steam, safety regulations were introduced to prevent future accidents. Periodic checks of the vessels by qualified engineers were a major element of this legislation. Insurers, together with the engineering profession, successfully developed control bodies and procedures to check pressure vessels, greatly improving the safety of technology and reducing the number of explosions of pressure vessels. 4. Nation states competing with insurance 4.1 Flood risks Many flood risks are linked to buildings erected, often illegally, in flood plains or at locations known to have a high flood risk. As land in flood plains is cheap, poor people often build their houses in flood plains, but cannot afford flood insurance, as demonstrated by the big Mississippi-Missouri flood in the 1990s. After a flood, losses to insured property are normally reimbursed by the insurance company involved, often taking into account a deductible. As major floods are almost inevitably declared disaster areas by governments, damages to non-insured property are fully reimbursed by government emergency funds, without a deductible or a premium. This raises the question of mandatory flood insurance for all house owners, even for those with a minimal flood risk. The advantages of mandatory flood insurance with a uniform common tariff are the speed in handling claims, the larger funds available, fairness in the treatment of losses, and reduced fraud. The disadvantages are the lack of a market signal for loss prevention (flood protection), the lack of incentives for political action to prevent repeated floods, and the exclusion of those people who may not insure even under a mandatory regime. Land planning, market structures, government aid, and regulation and insurance all interact in flood insurance. After the Mississippi-Missouri flood, Vice-President Al Gore suggested that uninsured victims in flood plains should only be reimbursed if they permanently gave up their existing homestead an approach that met with a lot of resistance. 4.2 Nuclear power stations In the 20 th century, third-party nuclear risks were not insurable, as no maximum loss could be put on a potential disaster. As countries had an interest in developing nuclear power stations, not least because they produce the raw material for bombs, an international Convention was agreed that made utilities exploiting nuclear installations legally liable for only the first U.S.$ 200 million of any loss. Furthermore, liability was limited to damages caused on the national territory. Today, tools of alternative risk transfer ( ART ), in the form of catastrophe bonds, could possibly cover a very limited number of nuclear power stations. Nuclear insurance is now limited by financial capacity, not technological risk assessment. In the case of nuclear energy, insurability is only an expression of economic value: in December 1999, 5 per cent of the shares of a nuclear power station located in Leibstadt, Switzerland, were sold by the then owner to a German investor for the price of A 60 million. But it was not the buyer who paid the A 60 million, but the seller, the price of getting rid of the
6 THE ROLE OF INSURABILITY AND INSURANCE 379 liability involved in owning the shares. This puts the market value of the nuclear power station at minus A 1,200 million! Due to the low cap on liability, nuclear energy is still heavily subsidized, as insurance premiums do not fully reflect the operational risks, and ignore the costs of other risks such as waste storage. Under the rules of the concept of insurability, it is probable that the energy market would have developed alternative non-carbon energies, such as solar or wind energy, at a much earlier stage. This case also questions the role of caps by nation states in turning uninsurable risks into insurable ones. A similar case is terrorism insurance for airlines after 11 September War risks, terrorism Thirty years ago, terrorism cover was often included in property insurance. But the events of 11 September have shown that people, rather than nature, pose the biggest risks, and that it is necessary to consider the maximum possible imaginable loss, not just the maximum possible loss. Terrorism risks have logically become uninsurable. In addition, 11 September has underlined the need for insurance to act in quantifiable risk fields, and for the state to take over the role of insurer of last resort in times of highest uncertainty. Insurance cover for airplanes has, in most cases, not been withdrawn, but premiums have dramatically increased to reflect the unprecedented losses caused by the four hijacked aeroplanes. Countries stepped in as guarantors to keep airlines from going into bankruptcy and to keep ticket prices at present levels. Politics have overtaken risk assessment by insurers as a guiding principle. 4.4 Reporting incidents and near misses A number of industrial sectors are obliged to report incidents and near misses. This obligation can be imposed by companies themselves (e.g. airlines), sectoral organizations (the U.K. chemical industry) or government organizations (Federal Aircraft Authority). The philosophy behind incident reporting is that most accidents have been preceded by a number of similar incidents without fatal consequences, which did not come to light because of reticence on the part of the staff involved. The chemical industry in the U.K. introduced incident reporting after the Flixborough accident. Reports are sent to a central body, which neutralizes the information by removing names and locations, and then distributes it to all its members. The system specifies that nobody will suffer any sanctions or punishment for faults committed. British Airways started its system after the crash of an airliner which ran out of fuel while flying a waiting loop. It was then discovered that a relatively high number of BA aircraft land with less than the minimal amount of fuel required, or have to ask for emergency landing due to lack of fuel. In most cases, these facts were not reported after landing. Furthermore, this situation often arose when a senior experienced pilot was in command. The younger co-pilot was aware of the situation but did not dare to point out the problem to the pilot. In some cases, the co-pilot needed psychiatric care after landing, which led to the fact being investigated. Regulations that punish human faults and errors often lead to the covering up of near misses, and thus waste the learning potential for loss prevention. Insurance does not have this effect, as it does not punish or sanction the faulty individual in the case of a near miss. If near misses are not reported, individuals tend to forget them or to integrate them as irrelevant or, worse, as positive experiences (e.g. drivers jumping red lights).
7 380 STAHEL 5. Nation states acting to prevent losses 5.1 Finland and Sweden: vision zero for casualties in road accidents Parliaments in both countries have made the vision of zero casualties in road accidents a clear national objective. In order to achieve this objective, a holistic approach is necessary, combining road design, car utilization and road user behaviour. Finland has reduced casualties by 50 per cent in the first five-year period, and again in the second period. Sweden has achieved the objective of zero casualties with regard to children within the first five-year period ( ). This strategy can only be chosen and imposed by nation states; it does affect road design but not automobile technology. Insurance will adapt the premiums of automobile insurance if the policy leads to a reduction in the amount of claims submitted, but otherwise can hardly play an active role. 5.2 The Swiss catastrophic loss register After the catastrophic fire of a chemical warehouse in Schweizerhalle, the Swiss government instructed the regional authorities of the cantons to list the potential disasters of a size equal to or bigger than Schweizerhalle. To everyone s general astonishment, the cantonal lists added up to 6,000 risks, a figure too high to be politically acceptable. So the list of potential disasters was scrutinized to identify catastrophic events, and develop methods of prevention or at least mitigation. One of the major risks consisted of two trains loaded with vinyl-chlorid ( VC ), travelling every week from Belgium to Italy through Lausanne and the Simplon rail tunnel. VC is a liquefied toxic gas heavier than air. If one of these trains had derailed in a station situated above a town, and the wagons been damaged, the population living below the tracks would have had to be evacuated. In the case of Lausanne, this would have meant evacuating about 60,000 people in less than 15 minutes an impossible feat. The only sustainable solution therefore was to eliminate the risk by convincing the Belgian VC manufacturer to change its production, i.e. to build new plants to produce and polymerise VC in both Belgium and Italy, with a corresponding loss in economies of scale. 6. Industry acting to prevent losses 6.1 Double hull crude oil carriers Conoco, then a fully owned subsidiary of DuPont de Nemours, was the first company to use exclusively crude carriers (ships) with double hulls. This was due to DuPont s corporate strategy of risk management, originally a manufacturer of explosives, with the objectives of zero incidents, zero deaths, zero losses. The additional construction costs were not offset by reductions in insurance premiums, but DuPont estimated that the prevention of a single major accident would pay for the additional cost in a longer-term perspective, taking into account the phenomenon called Accident Iceberg (a model suggesting that the large majority of damages such as loss of reputation are not covered by insurance). In the meantime, the International Maritime Organization (IMO) and the European Union increased political pressure to make double-hull crude carriers mandatory in the waters
8 THE ROLE OF INSURABILITY AND INSURANCE 381 of the E.U. This may, however, be challenged in court as it discriminates against some shipowners. 7. Conclusions The INTEREST project has shown that little is known outside insurance about the role of the concept of insurability as a tool to identify sustainable and environmentally benign technologies, at the crossroads of insurers and policymakers. Policymakers are only slowly discovering the concept of insurability as a potential borderline between technological innovation that needs policy intervention, and technological innovation that could more efficiently be dealt with by the marketplace through true costs of risk. The insurance sector has maybe done too little to date to put into the limelight its capacity to provide methods of identifying the insurable technologies among the emerging ones, and to draw policymakers attention to the factors why other emerging technologies are uninsurable, or to point out under which conditions these could become insurable. This is also a challenge for insurance to build a body of knowledge on how different actors have succeeded in shifting choices of technology towards loss prevention and sustainability. The present listing is meant as an incentive to collect the information necessary to establish such a body of knowledge. The conclusions of the INTEREST project suggest that the concept of insurability and its significance for actors beyond the insurance sector, and especially in the area of policymaking, has been underestimated in the past. Its impact could be greatly improved by a better dialogue between the actors involved in risk issues from the different fields of influence science and technology experts, policymakers, risk managers and insurers.