2.3 Objetivo general
2.3.1 Objetivos específicos
Turning to the failure inherent in the provision of natural disaster insurance, a key aspect of socio-economic coastal resilience is the ability to recover from a natural disaster through insurance (Berz & Smolka 1988, Doornkamp 1995, Clark 1998).
There are, however, various potential sources of natural disaster insurance failure, both on the demand-side and the supply-side.
Demand-side failures
In the case of Exmouth, those affected on the demand side would usually be property owners. Property owners obtain financial protection by transferring this risk through insurance, which also meets emotional needs, such as reduced anxiety, avoidance of regret and assurance of compensation in case a loss occurs (Kunreuther & Michel-Kerjan 2009). Possible demand-side failures created by disaster risk are summarised in Table 6-3. As can be seen in the table, the demand-side of insurance can fail as a result of five types of market failures. The first is moral hazard, which is a situation where insured or uninsured individuals raise the costs for the insurer, or relief agency, through risky behaviour, as they do not bear full costs for the risk (Arrow 1965, Baker 1996).
Government compensation can be affected by politics, as in the US, for example, where greater disaster compensation has been provided during presidential election years (Reeves 2004, 2005). Individuals may not have full information on the level of their exposure to risk, a failure situation known as asymmetric information, where insufficient or imbalanced information between contract holders can result in market inefficiencies (Stigler 1961, Akerlof 1970, Spence 1973). Failure can also result from cognitive bias, where even in cases where information is available, individuals may not obtain insurance or mitigate in proportion to risk exposure. Such a situation is described under prospect theory, where individuals diverge from optimising utility, which is a key assumption in neoclassical economics, because of underlying psychological traits and other motivations (Becker 1968, Kahneman & Tversky 1979).
Table 6-3 Summary of demand-side failures of natural disaster insurance compiled from relevant literature
Ideal conditions* Types of market failure and examples
Individuals take full responsibility to protect themselves from natural hazard risk
Moral hazard, charity hazard or natural disaster syndrome
Compensation from the government/donors reduces incentive to insure or mitigate (Kunreuther 1996, Kelly & Kleffner 2003, Raschky & Weck-Hannemann 2007) Little/no risk mitigation in California even following major earthquake damages (Kunreuther 1978, Palm et al. 1990) with similar findings with flood-proofing in other parts of the US (Burby et al. 1988, Laska 1991)
Full knowledge
Confusion because of expert disagreement (Bernknopf et al. 2006, Crichton 2008) Local governments withhold information because of concerns about lowered effect on property prices or getting sued for liability. Insurers may also withhold information to avoid disagreements on premiums rates (Gares 2002)
Full provision of
Incorrect interpretation and processing of risk information (Brilly & Polic 2005, Bouwer et al. 2007, Botzen et al. 2009a) or people ignore /do not understand risk probabilities (Kunreuther 1978, Palm et al. 1990)
Individuals are not able to connect the level of risk probability with the level of the insurance premium rate (Kunreuther et al. 1985, Kleindorfer & Kunreuther 1999, Camerer & Kunreuther 1989, Kunreuther et al. 2001, Huber 2004, Huber et al. 1997, Kunreuther 2006, Magat et al. 1987)
Unable to understand or distinguish between various probabilities of occurrence. For example in the Netherlands, more risk-tolerant people discounted the level of risk from that associated with damage from a 1in100 year flood ARI to that from 1in -100 year event, diminishing the importance of insurance (Kunreuther et al. 1985, Magat et al. 1987, Camerer & Kunreuther 1989, Huber et al. 1997, Kunreuther et al.
2001, Huber 2004, Brilly & Polic 2005, Bernknopf et al. 2006, Botzen et al, 2009a) Differing perception and risk-taking behaviour affected by factors such as proximity to a river, a recent major event, having purchased insurance in the past, having more children and being more educated (Gares 2002, Botzen et al. 2009a)
Mitigation and insurance is placed at the end of a list of competing priorities when discretionary income is limited, as is the case with low-income earners, females and older people (Kunreuther 1978, Lewis & Nickerson 1989, Botzen et al. 2009a)
The occurrence of the event must be independent of the will of the insured
Adverse selection
Individuals are more aware of their risk than insurers and obtain more cover. For example, in the Netherlands, those living in areas not protected by dikes have a greater demand for insurance (Rothschild & Stiglitz 1976, Camrer & Kunreuther 1989, Michel-Kerjan & Kousky 2008)
People take decisions to invest in mitigation/insurance based on the short-term, even where there are long-term benefits (Meyer & Hutchinson 2001, Kunreuther 2006).
Policies are cancelled when there has not been a recent major event (Kunreuther &
Roth 1998)
*Ideal conditions denote assumptions under the Arrow-Borch model and Swiss Re (1998) criteria
Adverse selection is where an individual has more information on risk than the insurer, resulting in those with high risk purchasing more cover (Freeman &
Kunreuther 2003). Failures can also result from inconsistencies associated with the higher value people place on the short-term, in comparison to the long-term, as described under the theory of intertemporal choice behaviour (Ainslie 1975, Raineri
& Rachlin 1993, Sozou 1998, Frederick et al. 2002).
Supply-side failures
As with failures on the demand side, natural disaster insurance also fails on the supply side owing to violations of the Arrow-Borch assumptions, and Swiss Re (1998) criteria. Possible supply-side failures created by natural disaster risk and examples from the literature illustrating their impact on the insurance market, are summarised in Table 6-4. There are two main types of market failures that affect the supply of natural disaster insurance: lack of information or uncertainties in the prediction of natural disaster risk, and the correlation of risk. Climate change and its associated potential impacts on the magnitude and frequency of meteorological disaster risk compounds the problem, imposing greater uncertainties in the calculus of insurance in high-risk areas.
Table 6-4 Summary of supply-side failures of natural disaster insurance compiled from relevant literature
Ideal conditions* Types of market failure and examples
Determination of the
Additional costs of research into risk, and other similar costs can result in 30% higher premium rates (Kunreuther & Michel-Kerjan 2009)
Limited coverage or the inability to set actuarially fair premium rates when potential liability cannot be determined (Born & Viscusi 2006, Kunreuther & Michel-Kerjan 2008)
Higher premiums charged when insurers disagree on the level of risk (Cabantous 2007) Natural catastrophes are not predictable on a year by year basis (Born & Viscusi 2006) Uncertainties prevent prediction based on the past, resulting in higher than optimal premium rates to cover unanticipated losses (Ellsberg 1961, Cummins 2006) Uncertainties, as to whether all measures to reduce risk have been taken, resulting in higher premiums, even for individuals who use mitigation measures (Aakre et al. 2010) Financial insolvency after major events, where insurers are unable to raise capital for large payouts or to raise capital quickly, resulting in exit from the market (Kleffner &
Doherty 1996, Cummins et al. 2002, Cummins 2006)
Following hurricane Katrina in the US, Poe Financial went bankrupt, Allstate Insurance lost US$ 38 billion in insured losses and exited several coastal states, and StateFarm opted not to renew some policies in the affected area (Born & Viscusi 2006) After hurricane Andrew in the US in1994, insurers suffered major underwriting losses, resulting in a restricted supply of insurance and higher premiums (Cummins 2006) Major events in the 1990s in the US destabilised the industry, and resulted in the failure of around 140 insurers (Cummins et al. 2002)
Uncertainty can lead to insurance premiums being set either too high or too low, which sends the wrong signal, resulting in underinsurance if premiums are too high, or unsustainable losses to insurers if premiums are too low (Born 2001, Gollier 2005) Additional uncertainties regarding climate change impacts on natural disaster risk and the lack of empirical analysis on how this will affect the industry (Born & Viscusi 2006)
Insolvency resulting from investing in risky financial markets (Born 2001) The insurer’s ability to provide adequate capital at given times is based on the underwriting cycle of the insurance industry (Cummins 2006)
Risks among several
Insurance works best with frequent events that do not cause severe devastation, which are statistically independent of each other and have a probability of being relatively evenly distributed over time (Born & Viscusi 2006). Several events over a short time period will impede the calculation of an economically viable premium, disrupting the market (Gollier 2005)
Economic inefficiencies can result in the industry holding on to large amounts of capital to cover massive damage resulting from low-frequency, very severe events (Jaffee & Russell 1997, Cummins 2006)
*Ideal conditions denote assumptions under the Arrow-Borch model and/or criteria required for the insurability of risk as defined by Swiss Re (1998) and Gollier (2002)
Uncertainties with regard to information on risk can create economic inefficiencies in the insurance market, including premiums rates being higher than actuarially fair. This can reduce the supply of insurance, and create withdrawal of coverage in areas where the level of risk is perceived to be too high or unpredictable (Born & Viscusi 2006, Cummins 2006, Kunreuther & Michel-Kerjan 2009). Correlated risk pertains to several individuals being affected at the same time by a major event, and all of these losses have to be covered by the insurer.
This is distinguished from other forms of insurance, such as that taken for automobiles, whereby a single accident will not usually affect a large number of insured individuals. As a result of correlated risk, the profit margins and financial viability of the insurer is reduced, resulting in a revaluation of the probability of loss in the future, higher premiums in the following years, and lower coverage as some insurers exit from the market. This often forces the government to step-in as insurer-of-last-resort (Freeman & Kunreuther 2003, Born & Viscusi 2006, Cummins 2006, Kunreuther & Michel-Kerjan 2009).