Paying the Price: The Status and Role of Insurance Against Natural Disasters in the United States (1998)

HOWARD KUNREUTHER¹

THE FINANCIAL COSTS OF natural disasters are on the societal radar screen. In the past they received relatively little notice except in the aftermath of a catastrophic event. The reason for our increasing awareness of the costs of natural catastrophes was succinctly summarized by two U.S. congressmen in an article in the Washington Post in 1995 entitled ''Natural Disasters: A Budget Time Bomb": "Over the past five years the cost of natural disasters has been rising at an alarming rate. In that time, 11 catastrophes have cost the nation more than $1 billion each. Hurricane Andrew and California's Northridge earthquake together cost more ($28 billion) than what the government spends annually on running the federal court system, aiding higher education and pollution control, combined" (Emerson and Stevens, 1995).

Even in a year when there are no significant single events, the trend is continuing, as indicated by the following quote in the Munich Reinsurance Company's summary of disaster losses in 1996: "1996 was a normal year for catastrophes, with losses of over

Houses in subdivision destroyed by Hurricane Andrew, 1992 (FEMA).

US$ 60 billion, of which US$ 9 billion were insured, and nearly 12,000 fatalities. However, the general trend toward ever-increasing numbers of catastrophes with ever-increasing costs is continuing" (Munich Re, 1997, p. 16).

As we will indicate in this book, in addition to an increase in the number of natural disasters, the dramatic rise in disaster losses has been caused primarily by a large increase in the population of hazard-prone areas, as well as a rise in the costs of construction. This book investigates the appropriate role of insurance to help stem this tide.

Insurance has an advantage over all other methods in the policy analyst's tool kit in that it rewards individuals prior to a disaster for investing in loss reduction measures through lower premiums, as well as paying these same people for damages suffered from a disaster. For insurance to be effective in both these roles, those who are at risk must bear a substantial portion of the costs of residing in hazard-prone areas. Otherwise they will have limited economic incentive to take protective actions and will rely on others to bail them out after the next disaster.

Insurance can play a key role in addressing two broad questions that are central to designing a hazard management program whose principal objective is to reduce future disaster losses:

• Who should bear the costs of making hazard-prone communities safer?

• Who should pay for the losses caused by disasters?

There are two criteria normally utilized in addressing these questions: efficiency and equity. By efficiency we mean the allocation of economic resources to maximize social welfare. Social welfare is defined by the citizenry and thus may vary from one political entity to another. A society that believes that every citizen has the responsibility to pay for the disaster losses of victims may find that taxation is the most efficient policy tool for generating the revenue to cover these costs. If, on the other hand, society believes each individual should be responsible for bearing his or her own burdens from natural hazards, then some form of insurance with variable rates based on the risks involved is likely to be viewed as an appropriate means of covering disaster costs.

Equity refers to concerns with fairness and the distribution of resources. An equitable distribution of resources may require the special treatment of certain individuals or groups at the expense of others. What may be viewed as equitable immediately after a disaster can be inefficient from a longer-term perspective if more people move into harm's way. For example, if uninsured disaster victims are guaranteed grants and low-interest loans that enable them to continue to locate their property in hazard-prone areas, and more people build in those areas, taxpayers will be subject to increasingly larger expenditures for bailing out victims of future disasters.

If private insurance is to play a central role in a hazard management program, as we feel it should, then the answers to the two questions posed above can be answered simply: Those in hazard-prone areas need to bear a substantial cost of making their communities safer and should be responsible for most of the losses after a disaster occurs. The larger the subsidy provided by the general taxpayer, the less important the role that private insurance can play in covering damage from disasters. In the long run, subsidies to special groups can only be successfully achieved and maintained by some form of social insurance. If there is genuine public concern with the increasing costs of natural disasters, as we believe there is today, then an insurance system with rates based on risk can serve as the cornerstone of a hazard management program.

This book assesses the role that private and public insurance has played in the past in dealing with natural disasters and suggests how it can play a more constructive role in the future. Our position is that the economic costs of natural disasters to the nation are too high and are likely to soar in the future unless some steps are taken to change recent trends. Insurers can address these problems in a constructive manner only through joint efforts with other stakeholders, and through the use

of strategies that combine insurance with monetary incentives, fines, tax credits, well-enforced building codes, and land use regulations. For example, one way to reduce future losses is to utilize insurance with well-enforced building codes and land use regulations.

The next section describes why natural disasters are of increasing concern to society, the nature of the available insurance, and the institutional arrangements that connect insurance to other concerned stakeholders. The section of this chapter entitled "Policy Implications" describes current public policy on natural disasters, the choices to be made by the public and private sectors in dealing with natural hazards, and the role that insurance can play in both reducing potential losses and providing compensation to disaster victims. The concluding section of this chapter describes the contents of this book.

An individual living in a hazard-prone area views a natural disaster as a low-probability, high-consequence event. Major hurricanes and earthquakes occur in specific regions, and then only infrequently, but when they do, the damage can be devastating, as evidenced by recent events such as Hurricane Andrew in 1992 and the Northridge earthquake in 1994. From the vantage point of a much large geographical area such as a state or the nation, the probability of some type of natural disaster increases significantly. Federal agencies that deal with disasters, such as the Federal Emergency Management Agency (FEMA), are almost certain to incur sizable expenditures in a given year, but cannot predict the total cost or how the payout will be allocated among various types of disasters.

Since 1989 insurance and reinsurance firms have suffered losses from disasters in the United States that have wreaked havoc with their balance sheets. Figure 1-1 depicts the magnitude of the catastrophic losses experienced by the insurance industry in the United States from 1949 to 1997 in 1997 dollars. The drastic change from 1989 to 1997 is obvious. Prior to Hurricane Hugo in 1989 (where insured losses were over $4 billion), the insurance industry had never suffered any loss of over $1 billion from a single disaster. Since that time 10 disasters have exceeded this amount in 1997 dollars (Gary Kearney, Property Claims Services, personal communication, 1998).

FIGURE 1-1 Insured catastrophe losses in the United States: 1949–1997 (in 1997 dollars). Source: Insurance Services Office, Inc.

Figure 1-1 sends a clear message to all concerned stakeholders that insured disaster losses have skyrocketed in recent years. The insurance industry, which was caught off guard by the very large increase in exposure in hazard-prone areas and hence had significantly underestimated the losses that could occur, is now concerned with significant natural hazards in certain parts of the United States. In these regions, especially Florida and California, insurers have prudently sought to substantially decrease their exposure and increase their rates. Some companies contend that they are facing an excessive risk of insolvency due to their high concentrations of catastrophe exposures and insufficient reinsurance (losses from Hurricane Andrew triggered the failure of nine small and medium insurers). Some also have expressed a concern that a fire sale of insurer assets (bond, stock portfolios, and real estate holdings) to pay claims from a large catastrophe could have severe repercussions for financial markets and further hamper the ability to recover. Others such as Cochrane, 1997, think that the indirect economic losses from a large-scale natural disaster would be relatively minor.

On a more positive note, new scientific studies, engineering analyses, and advances in information technology (IT) offer an opportunity to estimate the risks and potential losses of future disasters more accurately than in the past. More sophisticated risk assessments have reduced the

uncertainty associated with estimating the probabilities that earthquakes and hurricanes of different intensities and magnitudes will occur in specific regions. Engineering studies, building on the experience of past disasters, have provided new information on how structures perform under the stress of natural forces. The development of faster and more powerful computers enables one to combine these data in ways that were impossible even five years ago. These new developments should give a more complete picture of property at risk so that insurers can more accurately specify premiums that reflect their expected future losses.

As for the residents of hazard-prone areas, property owners today are experiencing greater difficulty in acquiring insurance and are paying considerably more for it than they have in the past. Moreover, these market conditions will become much worse if insurers continue to withdraw from the marketplace. Consumers have found it difficult to adapt to the rapid and severe changes in insurance market conditions following recent disasters. Overbuilding and high real estate prices in the Atlantic and Gulf Coast hurricane-prone areas have, until recently, been facilitated by relatively inexpensive and readily available property insurance. As shown in Chapter 3, the majority of home buyers also appear to have ignored catastrophe risks by failing to purchase earthquake and flood coverage. Commercial development has followed the population's movement to coastal areas, and this has increased the potential economic losses from natural disasters.

The resulting requests for significantly higher insurance rates and the desire to reduce writings threaten to deflate property values and hurt economies in high-risk areas. More specifically, many residents living in very hazard-prone areas may not be able to afford insurance premiums based on the actual risk. If forced to purchase coverage in order to maintain a mortgage, the residents may have to sell their houses. Purchasers will offer a lower price than in the past because of the higher insurance costs they will be forced to pay. Residents who are uninsured will request disaster assistance following the next catastrophic event. Chapter 9 proposes a hazard management program that addresses this issue.

This situation, which we term the natural disaster syndrome, poses an additional challenge to those concerned with the problem of reducing future losses. Lack of interest in protection against hazards prior to a large loss places a significant financial burden on society, property owners, the insurance industry, and municipal, state, and federal governments when these losses occur. More specifically, most homeowners, private businesses, and the public sector do not adopt cost-effective mitigation

measures to reduce potential losses from future disasters. A significant amount of damage would be averted if wind and seismic building codes were adopted and enforced, and if individuals took protective measures in advance of possible disasters. Serious consideration should also be given to reducing damage through more effective land use control measures.

This lack of interest in and enforcement of protective measures, coupled with the substantial growth of population in disaster-prone areas, has increased the probability that the losses will be severe when an earthquake, hurricane, or flood occurs. Given the large increase in the magnitude of losses from recent disasters, insurers and reinsurers are concerned about their financial ability to cover claims from future natural catastrophes.

The key interested parties concerned with natural disaster damage, their principal roles, and their linkages with each other are depicted in Figure 1-2. At the top of the figure are the reinsurance industry, capital markets, and federal government agencies, each of which has a special role to play with respect to providing protection against catastrophic losses. The Federal Emergency Management Agency (FEMA) and other federal agencies have stressed the importance of building codes and enforcement of regulations to reduce losses from natural disasters. Reinsurers relate to insurers in the same manner that insurers relate to property owners. They provide protection to primary insurers by insuring a portion of their claims in exchange for a premium. For all but the largest insurance companies, reinsurance is a prerequisite to offering insurance against natural disasters when there is a potential for catastrophic losses.

Recently the capital markets have provided private insurers access to funds in the form of catastrophic bonds. The insurer borrows from investors or an institution at higher than normal interest rates to cover extreme losses from hurricanes and earthquakes that exceed a trigger amount. If this amount is exceeded, then the interest on the bond, the principal, or both, are forgiven. For more details on these catastrophic bonds and other financing and hedging instruments, see Doherty (1997).

The primary insurance companies, as shown in Figure 1-2, provide direct insurance coverage to residential and commercial sectors for losses such as those caused by fires (including fires resulting from earthquakes) and those caused by wind damage from tornadoes and hurricanes. Primary

FIGURE 1-2 Role of interested parties concerned with natural disasters.

insurance companies offer this coverage through the standard homeowners' policies normally required as a condition for a mortgage, and through commercial multiperil policies.

Today coverage for shaking damage from earthquakes can be purchased as an addition to homeowners' policies in all states except California. A new public-private partnership, the California Earthquake Authority (CEA), formed in 1996, offers homeowners in the state earthquake coverage as a separate policy. The structure of the CEA and its impact on insurance protection of homeowners are discussed in Chapter 4.

As an alternative to a CEA policy, a few private insurance companies still offer earthquake insurance as a separate policy. The type of coverage and the premiums for this insurance vary among companies. For commercial structures, earthquake protection for property damage coverage is often included as part of a multiperil policy.

Since flood hazards are considered uninsurable by private insurers, FEMA provides direct insurance protection against flood losses through the National Flood Insurance Program (NFIP). The NFIP requires flood insurance and imposes hazard mitigation requirements on all properties with federally insured mortgages in flood-prone areas. The history of the NFIP and its current status are discussed in Chapter 6.

States have the principal regulatory authority over the primary insurance companies by virtue of the McCarran-Ferguson Act (Public Law 15, 15 U.S. Code, sec. 1011 et seq.) passed in 1945. Each state has an insurance official who has two primary areas of responsibility: (1) monitoring and overseeing the financial solvency of the insurers, and (2) examining insurers' rates and market practices. The National Association of Insurance Commissioners (NAIC), through its advisory recommendations, plays a key role in state regulators' efforts to coordinate and strengthen their oversight of the insurance industry (Klein, 1995).

"Property at Risk," in Figure 1-2, consists of businesses, homes, and rental units located in hazard-prone areas. Property owners and tenants can reduce their potential losses from future natural disasters by adopting mitigation measures, and can protect themselves financially against the consequences of future damage by purchasing insurance and renewing their policies over time. Several other interested parties play key roles in the design and enforcement of insurance and damage mitigation requirements: financial institutions, through specific requirements as a condition for a mortgage; the construction industry, by designing safer structures; state and local governments, through ordinances and building codes; and the real estate sector, through provision of information on hazards to potential buyers and owners.

A coherent strategy has not yet been developed for coping with the changing role that natural hazards are playing in our lives. There is general agreement that it is important to take steps to reduce losses from future disasters. In fact, FEMA introduced a National Mitigation Strategy in December 1995 with the objective of strengthening partnerships

between all levels of government and the private sector to ensure safer communities. This strategy was developed with input from state and local officials as well as individuals and organizations with expertise in hazard mitigation (FEMA, 1997).

A parallel effort has been undertaken by the Central U.S. Earthquake Consortium through a pilot project in two communities which emphasized the importance of local and business sector involvement in mitigation activities (Central U.S. Earthquake Consortium, 1997). Success in these endeavors requires the identification of cost-effective mitigation measures as well as increased expenditures to make structures safer in hazard-prone areas. To date, property owners have shown little inclination to incur such expenditures voluntarily, so there is a need for incentives, including insurance premium reductions, to encourage the adoption of these measures. In addition, well-enforced building codes are also needed to reduce collateral losses sustained in addition to those suffered by the property owner, such as damage to other structures and social costs arising from property damage (Kunreuther, 1997).

Current natural disaster policy places a large financial burden on all taxpayers after a disaster occurs. Under the Stafford Disaster Relief and Emergency Assistance Act of 1988 (P.L. 100-707, 42 U.S. Code sec. 5121 et seq.), the federal government provides funds to cover at least 75 percent of the costs of rehabilitating public facilities (U.S. Congress, 1995). For catastrophic events such as Hurricane Andrew in 1992 and the Mississippi floods of 1993, the federal government covered the entire cost of public facility repairs; it paid 90 percent of these costs after the Northridge earthquake, with the remainder financed by the State of California. Thus, it is not surprising that there has been little interest by most municipalities in investing in loss reduction measures for their facilities; city officials probably assume damage will be covered by federal or state funding. For the same reason, a number of local governments have shown little interest in adopting mitigation measures or purchasing insurance against losses to their buildings (Burby, 1992). A study by French and Rudholm (1990) of the damage to public property in the Whittier Narrows (California) earthquake of October 1987 revealed that few public buildings were protected by earthquake insurance, even though it was readily available from the private sector.

With respect to the private sector, the federal government offers assistance

to uninsured and underinsured disaster victims through a Small Business Administration (SBA) disaster loan program. Under the current arrangement, homeowners and businesses suffering damage from a disaster can obtain a low-interest loan to aid their recovery. The interest rate varies between 4 percent and 8 percent. These programs can be costly to taxpayers if many such loans are provided at below-market rates. During the period from 1977 through 1993, the SBA loaned approximately $21 billion to disaster victims (U.S. Congress, 1995).

One question that needs to be addressed concerns the appropriate roles of the private and public sectors in financing the cost of recovery from large-scale natural disasters. To the extent that private insurance markets provide protection against catastrophe risk, policymakers must decide how these markets should be regulated. They must also determine the role of land use regulations and building codes and the extent to which private choice and incentives will guide hazard mitigation efforts. Within the realm of public choice, decisions also must be made with respect to the delegation of authority among the different levels of government and its agencies. In evaluating these options policymakers must consider how various government actions affect the behavior of firms and individuals in responding to catastrophe risk.

The choice between public mechanisms and private markets for financing catastrophe risk depends on how specific measures affect the decision processes of the various interested parties as well as the final outcomes. For example, it is very difficult to enforce significant cross-subsidies between different groups of risks (if that is what policymakers want to achieve) through the regulation of private insurance markets. Private businesses operating in a competitive market have strong economic incentives to avoid selling their products and services at below cost. Over time they can usually find ways to avoid or minimize their volume of such business, or to withdraw from those markets entirely. In order to implement a policy whereby rates of low-risk policyholders are used to subsidize those of higher risk, it would very likely be necessary to have legislation providing that banks and financial institutions require individuals and firms to purchase catastrophe insurance as a condition for a mortgage, and for some government agency to provide this insurance.

There is also a need to coordinate government policies affecting disaster risk. Specifically, in analyzing government policies toward insurance

we must consider the impact of other private and public sector activities on hazard mitigation and the funding of recovery following a disaster. For example, the reliance on post-flood assistance and the failure of financial institutions to enforce flood insurance requirements for federally guaranteed home mortgages have adversely affected the National Flood Insurance Program's ability to encourage the purchase of flood insurance. Amendments to the NFIP in 1994 penalize banks if they do not enforce these insurance requirements. These amendments provide for a maximum of $350 for each violation, subject to a maximum calendar year amount of $100,000 for any single regulated lending institution or enterprise. As a result, there has been a substantial increase in the number of NFIP flood insurance policies purchased.

The principal problems related to natural disasters are the risk of large and uncertain economic losses, the lack of availability of adequate or desired insurance coverage for some risks, the high cost of coverage, and insurance company financial difficulties and potential insolvency. In a perfectly competitive market the price of coverage should reflect the degree of risk, with a normal markup for administrative costs. However, if some of the conditions for competitive markets are not present—such as adequate information, barrier-free entry and exit, and unrestricted flow of capital—the result could be failures in the private market.

Incomplete information and misperceptions about the risk of natural hazards could cause property owners to demand a suboptimal amount of insurance coverage. Uncertainty about the true risk of loss could cause insurers to refuse to supply coverage or charge very high prices for catastrophe insurance. If insurers fail to correctly distinguish between risk types, for whatever reason, informed individuals in high-risk areas will be more inclined to purchase coverage, while those in less hazard-prone areas will be disinclined to buy coverage, resulting in a skewing of the market and an unprofitable book of business for insurers.

Entry barriers can restrict the supply of insurance, decrease competition, and raise prices. Exit barriers can further discourage entry of new insurers. Restrictions on the flow of capital also can limit the supply of coverage in places where there is considerable demand for insurance protection. Government expenditures, taxes, and regulations can create perverse incentives which distort market decisions and cause inefficiency. The challenge is to determine how government intervention can diminish

or offset these market failures and increase economic efficiency and equity. The nature and effects of insurance market failures and government intervention are discussed further in the chapters to come.

Figure 1-3 provides a road map and guide to what follows in this book.

Chapter 2 focuses on the supply side of the market by examining the insurability conditions for natural disasters. Private insurers are reluctant to continue providing widespread coverage against earthquakes and hurricanes as they have in the past. New institutions such as the California Earthquake Authority and the Florida Joint Underwriting Authority were formed in these two states because of the severe losses from the Northridge earthquake and Hurricane Andrew, respectively. Chapter 2 also examines the roles of the reinsurance industry, state insurance pools, and the capital markets in offering protection against large disasters that threaten to reduce insurers' surpluses to unacceptably low levels.

Chapter 3 focuses on the behavior of those in need of protection—residents in hazard-prone regions—and on the changing demand for residential disaster insurance in the United States. Prior to 1990 residents in hazard-prone areas had limited interest in voluntarily purchasing such coverage. Following the spate of recent disasters, the demand for insurance protection increased, as evidenced by the number of residents purchasing earthquake insurance in certain parts of California following the Loma Prieta earthquake of 1989 and Northridge in 1994. One open question concerns how price sensitive consumers are in purchasing coverage. More specifically, what impact would a rise in earthquake, wind, or flood insurance prices in certain parts of the country have on voluntary demand for coverage?

The next three chapters examine the role of insurance and other policy instruments in relation to three different types of disasters. Chapter 4 provides a history of earthquake insurance, with a focus on California, the state most concerned about and affected by that hazard. Chapter 5 examines the challenges insurers face in providing coverage against wind damage from hurricanes, using Florida as a case study. Chapter 6 reviews the changes in the National Flood Insurance Program since its establishment as a federal program in 1968 and describes its current status.

Chapter 7 examines the role that insurance and mitigation can jointly play in reducing future losses. It gives special emphasis to building codes

and their enforcement through the assistance of financial institutions and the building industry. Chapter 8 is devoted to the appropriate roles of the state insurance regulator with respect to setting rates, imposing coverage requirements, and minimizing the probability that firms will become insolvent following a catastrophic disaster.

Chapter 9 proposes a hazard management program for reducing disaster losses and providing money for recovery. The success of such an effort depends on the ability of insurers to work closely with other interested parties such as financial institutions, public sector agencies, and the capital market. This concluding chapter explores the potential role of new financial instruments, as well as some type of federal reinsurance to cover a portion of insurer and reinsurer losses from a very large catastrophe. At the end of each section of the final chapter we raise a set of open issues and questions for future research.

The three appendixes supplement the book with information on the availability of commercial insurance for natural disasters (Appendix A), the state of the art of modeling the risks from natural hazards (Appendix B), and links to information on the World Wide Web (Appendix C). These Internet linkages will enable the reader to stay up-to-date on current activities by interfacing with the many sites that are listed. A glossary at the end of the book defines technical and insurance-related terms for the general reader.