The previous chapter explained why donors and IFIs should intervene in catastrophe insurance markets in developing countries, but these reasons are not sufficient for donor-supported public intervention. Such involve- ment comes at a cost; therefore it should be demonstrated that interven- tion will result in an improvement and that these benefits will exceed the costs. There is a risk that public intervention may be ineffective (that is, no efficiency gains are achieved) or even detrimental (that is, there are efficiency losses).
There are two main risks of public intervention in catastrophe insur- ance market. First, the government may want to develop public catas- trophe insurance programs. As shown below, government is unlikely to implement the basic principles of efficient catastrophe insurance, leading to a financially unsustainable program that risks crowding out the pri- vate sector. Second, the government may be tempted to increase market penetration by offering insurance premium subsidies to lower the cost of insurance. Premium subsidies may distort the price signal and thus lead policyholders to under-invest in risk-mitigation activities or to invest in nonviable activities.
Public Catastrophe Insurance Programs
Establishment of efficient catastrophe insurance programs depends on the integration of four key components: risk assessment, risk pooling, risk segregation, and control of moral hazard (see Box 3.1). However, public catastrophe insurance programs run by governments are unlikely to fol- low these key principles, thus leading to unsustainable solutions.
Risk assessment is critical in enabling insurers to discover the true costs of risk. Sophisticated risk assessments inform insurers about the
Principles for Public Intervention in the Catastrophe Insurance Markets 77
policyholders’ risk characteristics and help to control adverse selection by segregating or classifying heterogeneous risks into more homogeneous cat- egories. However, segregation is sometimes viewed as socially unacceptable, because it may not be compatible with social and solidarity objectives. As a consequence, public insurance programs are likely to under-invest in con- trolling adverse selection through risk classification and to offer the same average premium rate to all policyholders. This leaves little incentive for
Box 3.1 Basic Principles for Efficient Catastrophe Insurance Risk assessment. Risk assessment is used to discover the true underlying
cost of risk, which is pivotal to designing cost-effective catastrophe insurance programs. Risk assessment requires data collection and the application of statistical and actuarial modeling techniques. Catastrophe risk models can also be used to assess the risk of loss from disasters and improve government ability to mitigate the losses (see Annex 3).
Risk pooling. Insurance reduces the risk level to society by aggregat-
ing non-correlated risks. This makes the average loss more predictable and distributes the costs more efficiently. However, because pooling does not fully eliminate risk, insurers must maintain equity capital to absorb any variation around the expected loss. The amount of capital required is based on the independence of the risks; those risks that are correlated require more risk capital set aside than independent risks. The insurance industry has developed several techniques to make co-variant risk insur- able: risk diversification over time, bundling catastrophe risk with other property risks (for example, fire), reinsurance, and so on.
Risk classification. Through appropriate data-collection and risk-
underwriting techniques, insurers attempt to distinguish relatively high risk from low-risk applicants and then assign policyholders insurance premiums that accurately reflect their underlying risk characteristics. Insurance performs this function by segregating risks into the appro- priate risk pools, that is, through risk classification. Risk classification and risk-based pricing reduce the chance of adverse selection, which occurs when high-risk policyholders over-consume insurance because their premiums are subsidized at pooled insurance rates. Charging risk- based premiums provides a market signal to policyholders about the
those with the greatest risk to lessen their exposure, while at the same time, makes insurance overly expensive for those with lower exposure.
Even if premium rates are appropriate on average, government- provided or government-backed insurance may provide cross-subsidies among classes of buyers, to promote equal treatment of all participants. For example, governments often undercharge buyers with high expected loss costs and overcharge buyers with low expected loss costs, because it is often politically unfeasible to charge citizens different rates. Providing insurance at sub-competitive prices or subsidizing high-cost insurance buyers creates moral hazard problems whereby policyholders under- invest in loss prevention or over-invest in high-risk activities or geo- graphical areas.
Insurers control adverse selection through various coverage-design feature, including deductibles, coinsurance, and policy exclusions. How- ever, such coverage limitations may be viewed as inconsistent with the government’s objective of offering broad coverage. For example, limiting
Box 3.1 (Continued)
cost of engaging in activities that generate the risk. This gives policy- holders incentives to invest in risk mitigation to lower their direct risk, and thus, their insurance premium (so long as any risk reduction is reflected in the insurance premium).
Control of moral hazard. The insurance industry has adopted a
number of measures to reduce the impact of moral hazard, which occurs when policyholders change their behavior because the risk of loss is borne by someone else. To prevent moral hazard, insurers strive to align the policyholder’s interests with the overall interests of the insurance pool. Commonly used tools to prevent moral hazard are deductibles and coinsurance, included as part of the insurance contract. Deductibles and coinsurance give policyholders the incen- tive to operate safely and engage in risk mitigation to avoid paying their portion of the loss. A final tool is the exclusion of some events from insurance coverage.
Principles for Public Intervention in the Catastrophe Insurance Markets 79
access of coverage by refusing coverage to citizens most at risk to disasters, or charging much higher premiums to reflect greater risks, may not be politically viable to a government attempting to establish catas- trophe insurance coverage.
Under voluntary insurance, the failure to charge premiums that accurately reflect the underlying risks, based on government pressure to promote inclusion of all high-risk individuals, leads to a vicious cir- cle of adverse selection. If the insurance industry is required (by law or decree) to charge equal rates to policy holders, this results in wealth redistribution from low-risk to high-risk policyholders, because low- risk policyholders are overcharged and high-risk policyholders are undercharged (subsidized).
To increase penetration of catastrophe insurance and depth of the market, governments may be tempted to require certain segments of the population to purchase coverage. Compulsory insurance may be viewed as a solution to adverse selection because it forces low-risk policyholders to remain in the insurance pool. It is also related to the principle of sol- idarity among citizens. Nevertheless, compulsory insurance may lead to a social sub-optimum because part of the population is forced to pur- chase catastrophe insurance.
However, because social or political constraints can prevent govern- ments from controlling adverse selection through the application of sound insurance management practices, governments usually are ineffective in providing primary insurance coverage.
The efficient financing of natural disasters should thus rely on a pub- lic-private partnership between the private insurance, reinsurance indus- tries, and governments. Figure 3.1 summarizes the public and private financing responsibilities. Governments, eventually backed by donors, should focus on:
• Stimulation of competitive private insurance markets (through the development of risk-market infrastructures and the financing of top layers of catastrophe risks);
• Provision of post-disaster loans and grants to the poor and disadvan- taged; and
Insurance Premium Subsidies
Instead of actively managing catastrophe insurance programs, govern- ments often choose to intervene in the insurance market to promote market development and penetration through financial incentives. Governments frequently subsidize catastrophe insurance, and particu- larly agricultural insurance products. Regardless of the form, govern- ment subsidies are usually designed to increase insurance penetration by lowering insurance premiums charged to the policyholders. Such public subsidies may be justified by the existence of market imperfec- tions, but there is a risk that public intervention distorts the price sig- nal and crowds out the private sector.
In a well-functioning private insurance market, premiums should be risk-based and differentiated so that each buyer pays a premium sufficient to cover his or her own expected loss and expense costs as well as a profit loading to compensate the insurer for bearing insurance risks. With risk- based premiums, buyers bear the full costs of their risk-generating activi- ties and thus have incentives to engage in risk mitigation and not to overindulge in risky activities. Subsidized catastrophe insurance induces overinvestment in structures in risky areas such as flood plains and coastal communities exposed to hurricanes. It also induces subsidized policyhold- ers to buy more insurance than they would purchase under risk-based
Source: Authors, from Gurenko and Lester (2004).
Figure 3.1 Public-Private Partnership in Catastrophe Risk Financing
Government Donors/IFIs Post-disaster subsidized loans and grant facility Lifeline infrastructure The poor and disadvantaged Private reinsurance/capital markets Catastrophe (re)insurance pool
Domestic insurers Propertylenders Small businesses Commercial farmers Middle class housing
Donors/IFIs
Principles for Public Intervention in the Catastrophe Insurance Markets 81
insurance rates (Grace, Klein, and Kleindorfer 2004). These adverse-incentive effects increase the expected losses from catastrophes and impose costs on governments, taxpayers, and donors.
Two main types of insurance subsidies can be defined. Market-enhancing
insurance subsidies are subsidies that enable or promote competitive
insurance markets. These subsidies focus on the development of public goods and technical assistance that enhance the risk market infrastructure, such as data collection and management systems, catastrophe risk models and legal and regulatory framework. Social insurance premium subsidies are provided by governments as part of social safety net programs. The most common form of social premium subsidy is direct-premium subsidies that are proportional to the insurance premium that would be charged in a private insurance market.
Market-enhancing insurance subsidies aim to support a healthy and
sustainable competition among insurance and reinsurance companies by reducing the startup costs and entry barriers. Such subsidies ultimately con- tribute to the reduction of the insurance premiums and therefore benefit the policyholders. Programs can be designed to address the market imperfec- tions identified in the previous chapter, while avoiding market disincentives that lead to over-consumption of risky activities and higher loss costs.
Favorable subsidy programs include the provision of public goods that contribute to the development of insurance market infrastructure, such as data collection and management, development of catastrophe risk mod- els, the creation of legal and regulatory frameworks, and so on. Other favorable subsidy programs include those that build the capacity of the domestic insurance industry and facilitate the transfer of risk to global (re)insurance and capital markets. Effective programs can also be created to educate the public about the benefits of insurance and to increase insur- ance literacy. Providing technical assistance and development aid to the domestic insurance industry is another form of subsidy that can improve the ability of domestic markets and institutions to finance catastrophe losses without leading to harmful disincentives.
The Turkish Catastrophe Insurance Pool is an illustration of a catastro- phe insurance program with risk-based premium rates without public subsi- dies. Premium rates differ by risk areas (5 zones) and by type of construction (3 types). Likewise, for the Caribbean Catastrophe Risk Insurance Facility,
which provides sovereign insurance to Caribbean governments, the insur- ance premium reflects country-specific risk exposure. In agriculture, the live- stock index-based insurance program in Mongolia and weather-based crop insurance program in Malawi provide insurance coverage where the herders and the farmers pay an actuarially fair premium without direct premium subsidies (see Annex 5). Most of the hail-insurance programs for crops are commercially viable without direct premium subsidies.
Social insurance premium subsidies tend to have highly distortional
implications for the insurance markets and risk management behavior of the policyholders. They should be avoided and governments should offer separate safety net programs managed by the public sector. An example of a social insurance premium subsidy program is a scenario where 50 percent of the risk-based premium may be paid by the government, and the other 50 percent is paid by the policyholder. Experience shows that this form of premium subsidy is usually inefficient and increasingly expensive because direct premium subsidies tend to: i) be untargeted and available to all policyholders, whatever their ability to pay, because it is politically difficult to discriminate regarding the level of premium subsi- dies among the population; ii) be permanent, even though the govern- ment initially introduces them as temporary subsidies; iii) represent an increasing fiscal burden for the government, because the eligibility crite- ria are relaxed or the subsidy levels increase; and iv) mainly benefit pol- icyholders located in high-risk zones.
The following two examples demonstrate the imperfections of social insurance premium subsidies. Proportional insurance-premium subsidy programs are widely used in crop insurance. In 2006, U.S. crop insur- ance premium subsidies were US$2.3 billion (60 percent of total crop insurance premiums),1meaning that farmers paid on average 40 percent
of the total insurance cost. These subsidy programs mainly benefit large farmers (who insure a large value at risk) and farmers growing high-risk crops (whose risk-based insurance premium rate is higher than that for farmers with low-risk crops). In addition, these programs are largely captured by insurance companies. The GAO (2007) reports that during 2002–2006, U.S. Department of Agriculture crop-insurance subsidies resulted in underwriting profits of US$2.8 billion for the insurance industry. These gains represent an average annual rate of return of
Principles for Public Intervention in the Catastrophe Insurance Markets 83
17.8 percent over this 5-year period, approximately three times the rate of underwriting return on private market property casualty insurance during this period.
Another illustration of social insurance premium subsidies is India’s National Agricultural Insurance Scheme (NAIS), a government-sponsored crop insurance program that protects Indian farmers against the adverse weather events and pest infestation. NAIS is heavily subsidized by the Indian country and state government and is designed as a social scheme, with limited risk classification.2 Rates are capped at 3 percent for food
crops nationwide—well below the actuarially sound premium rates. The lack of risk classification and premium differentiation does not give farm- ers incentives to change their farming practices or to shift to more viable crops. For example, groundnut farmers in states heavily exposed to drought risk should pay premium rates of more than 30 percent under actuarial pricing (World Bank 2007c) instead of the 3 percent cap legis- lated by the government. Since they pay only a 3 percent premium rate under NAIS, they have no incentive to invest in risk mitigation or to shift toward more drought-resistant activities. Although it is compulsory for borrowing farmers, this scheme is voluntary for non-borrowing farmers. As one would expect, the only non-borrowing participants in the NAIS are high-risk farmers because of a lack of price discrimination that allows them to receive coverage at a cost below the market price. As a result, the long-term average loss cost (the ratio of losses to the insured value) is 26.4 percent for non-borrowers, compared to 9.9 percent among bor- rowing farmers. Implicit premium subsidies represent approximately 75 percent of the actuarially sound premium rates.
Social insurance premium subsidies tend to have highly distortional implications for the insurance markets and risk management behavior of the policyholders. As such, they should be avoided and govern- ments should offer separate safety-net programs managed by the pub- lic sector. However, when the public financial delivery systems face severe leakages or even corruption, the insurance industry delivery systems may be more efficient in providing assistance to poor house- holds. In this case, social premium subsidies targeted to poor house- holds may be justified, as part of a social safety net program rather than an insurance program.