CIAO DATE: 12/03

Government Reinsurance of Terrorism Coverage

Peter J. Wallison

On The Issues

November 2001

American Enterprise Institute for Public Policy Research

Property and business insurers have threatened to discontinue terrorism coverage unless the federal government assumes substantial portions of the risk. The best program for doing so would be temporary, would expose the insurance industry to substantial first-dollar losses, and would not regulate premium rates or charge insurance companies a premium for the risk the government assumes.

Estimates of insured property and business losses exceeding $50 billion in the destruction of the World Trade Center, and uncertainty concerning the future course of terrorism in the United States, have caused reinsurers around the world to threaten withdrawal of terrorism coverage. As a result, primary property and casualty insurance companies have indicated that, unless the government steps in to provide last resort reinsurance coverage, they will seek to exclude terrorism risk from new and renewal insurance contracts beginning in 2002. Such exclusions could create a significant obstacle to the nation's economic recovery. For this reason, the administration, the insurance industry, and members of Congress are considering plans for the federal government to assume substantial portions of terrorism insurance risk—and thus encourage the inclusion of terrorism coverage in insurance contracts coming into force in 2002 and future years.

It is certainly plausible that the withdrawal of terrorism coverage—at the very moment when the risk of terrorist attacks has become palpable—could have serious, economy-wide costs. Banks that finance buildings and other facilities are unlikely to lend for new construction where there is a risk that the collateral will be destroyed with no insurance behind it. It is possible, moreover, that existing loans might be deemed to be in default if insurance coverage—because of a new terrorism exclusion—is considered inadequate. Widespread denial of financing for new construction, or the defaulting of loans to weak borrowers, would certainly impede the economy's ability to snap back from the long slowdown it has been experiencing.

But is it plausible that the reinsurance industry will deny terrorism coverage? The insurance and reinsurance industries are highly competitive. Primary insurers, which offer coverage to individual firms, can lay off a portion of their aggregate risks with a large number of reinsurers. There is even a tertiary market in which reinsurers can lay off some of the risks they have assumed from the primary insurers. In these highly developed, global markets, casualty loss risks—including risks of enormous losses from hurricanes, earthquakes, and other catastrophes—are routinely sliced and diced through a myriad of private contracts, the effectiveness of which can be judged by the fact that the huge losses of September 11 will be spread throughout the world. When all of those losses have been accounted for, it will almost certainly turn out that no major primary insurer involved with the World Trade Center will become capital-impaired.

In these circumstances, one might assume that at least some reinsurers, who are in the business of taking and spreading risks in exchange for premiums, would be willing to take on the newly evident risks associated with terrorism. Nevertheless, the insurance industry is virtually unanimous that there will be no terrorism coverage after January 2002, and there are two good reasons for thinking the industry is correct.

First, although the industry takes risks on events that cannot, by definition, be predicted with any certainty—such as earthquakes, fires, hurricanes, and floods—it prices them according to a long history of these and other catastrophes that permits some estimation of potential losses over time. But terrorism at least on the scale of the World Trade Center disaster has no precedents. That a large number of sophisticated, highly trained individuals are willing to sacrifice their lives to destroy property and kill civilians is entirely new. Industry actuaries cannot predict either the likelihood of another occurrence or the size of the potential loss. Despite the fact that insurance and reinsurance companies are in the business of taking risks, these risks are to some extent limited by the rational estimates of actuaries. Without these estimates, insurance industry experts say, the risks simply cannot be priced.

Second, the losses arising from the catastrophe of September 11 may, when all is said and done, be very large in relation to the total capital of the reinsurance industry. Another such loss, occurring soon after the first, could place large numbers of companies in jeopardy of failure. What might have been an acceptable loss—even though large—before September 11 becomes unacceptable afterward if it amounts to betting the farm. The uncertainty associated with a potential loss is compounded by uncertainty about the actual size of the earlier loss. Under these circumstances, elementary prudence requires that no further incalculable obligations be taken on.

 

Designing a Reinsurance Program

These arguments are difficult to dismiss, and are likely what have convinced the White House, the Treasury Department, and ordinarily conservative, market-oriented legislators to support some kind of assistance to the insurance industry. The form that this assistance takes, however, raises several difficult and interrelated policy issues.

First, insurance is a business that takes risks in exchange for payment in the form of premiums. When risks turn into losses, premiums are raised to replenish the industry's capital and to provide a return to its shareholders. But higher premiums are also intended to prepare the industry for the higher losses that it has now discovered are associated with a particular risk. If the industry does not eventually assume the risk for which it is being compensated, the higher premiums are a windfall. Under these circumstances, any government program that allows the insurance industry to recoup its losses on September 11 must also be structured so that the private insurance industry eventually takes on full responsibility for terrorism losses. A program that allows the insurance industry to recoup its losses without eventually reassuming responsibility for terrorism coverage would be badly designed.

Second, one of the dangers associated with insurance of any type is moral hazard. This term refers to the fact that insurance reduces the incentives of an insured to protect itself against loss—since that loss will be borne by the insurer. For this reason, insurance companies by contract specify steps that an insured must take to reduce the likelihood of loss. Homeowners see this in contractual provisions that reduce premiums for homes with sprinkler systems or burglar alarms. In the commercial world, many more intrusive steps are specified in insurance arrangements, and these are salutary because they reduce the likelihood of loss while spreading any actual losses widely through insurance and reinsurance arrangements. Thus, any government program to cap the insurance industry's terrorism losses must not reduce the industry's incentives to prevent moral hazard by requiring insured parties to take appropriate security measures.

Third, if the risks of terrorism cannot be priced—the rationale for a government program in the first place—then premiums for insurance contracts that cover terrorism risk will have to be set somewhat arbitrarily. This seems to open the door to government regulation of insurance premiums. At the same time, the speed with which the insurance industry replenishes its capital—that is, the size of premium increases in the coming years—could have a significant impact on the nation's economic recovery, and will certainly generate complaints by insureds that they are being hit too hard by rate increases.

Congress will therefore be interested in reducing the impact of insurance premium increases—and its new role as an insurance underwriter will provide ample opportunity to pursue that interest, now or later on. It will be very important for Congress to resist this temptation. The commercial casualty insurance markets are highly competitive. There is good reason to believe that competition among primary insurers will drive rates down over time to levels that appropriately balance the interests of insurers to recapitalize and insureds to resist sharp premium increases. As long as the industry will eventually pick up all terrorism risk, higher premiums are a social cost that should be borne. In addition, higher premiums will mobilize additional capital through catastrophe bonds and other mechanisms through which private investors become de facto insurers of particular risks.

Fourth, the government has an interest in protecting taxpayers as well as encouraging economic growth, and here the two may be in conflict. In past government programs to assist the private sector—the Chrysler bailout in the late 1970s, the Continental Illinois bailout in the 1980s, and the recent airline relief legislation—the government has exacted compensation for allowing otherwise failing companies to survive on government credit. Fannie Mae and Freddie Mac notwithstanding, the government generally receives compensation—a guarantee fee—when its credit is used by the private sector. Following these examples in the case of public reinsurance, the government would charge the primary insurance companies some kind of premium for the terrorism risk the government is assuming.

There are, however, good reasons not to compensate the government. It is the government's responsibility to protect the nation against terrorism, and its failure to do so would be the proximate cause of the losses to the insurance industry. In a sense, paying premiums to the government for the risk it is assuming on terrorism is like paying twice for the same service. In addition, and perhaps more important, casualty loss insurance coverage is so pervasive in the U.S. economy that a premium paid to the government would ultimately be passed through to virtually every business—amounting to a new tax at a time of economic weakness.

These circumstances suggest the following principles for a government reinsurance program:

• The insurance industry should be exposed to substantial first-dollar losses, much like a deductible in auto or health insurance, before the government reinsurance kicks in. That would discourage moral hazard while performing the essential public purpose—capping the industry's total potential liability at a time when that liability cannot be estimated and priced and when the industry's capital is weakened to an unknown extent.
• The ultimate objective of the program should be the eventual termination of government responsibility and the assumption of terrorism risk entirely by the private sector. Accordingly, over time, the industry's liability should increase—probably through taking increasing percentages of the government's reinsurance liability. This percentage increase can be adjusted with experience. If terrorism is in fact reduced or effectively eliminated by the current worldwide effort led by the United States, then the industry's loss-sharing with the government will accelerate; if the experience is otherwise, loss-sharing will not increase as quickly.
• Efforts to impose rate regulation as a "price" for government's absorbing some terrorism risk should be firmly resisted. Although there is some public interest in holding down rate increases, that interest is best served by getting the private sector back in the business of insuring against terrorism risk as quickly as possible—a task that rate regulation can only delay. Insurance rate regulation is a fool's errand in any event: States that regulate insurance rates almost always end up with higher rates (because of suppressed competition) than states that regulate lightly or not at all.
• The government should not charge a premium for taking terrorism risk. A new (de facto) tax and revenue source of this kind will create an incentive to keep the government reinsurance liability in existence long after it is needed; absence of compensation will provide an incentive for Congress to end the program as quickly as possible.

If Congress addresses the moral hazard problem, keeps the program temporary, and resists temptations to raise new government revenues or regulate private contracts, it will have provided an exemplary civics lesson—demonstrating the capacity of representative government to produce sensible policy in times of duress and uncertainty.

 

Peter J. Wallison is a resident fellow at AEI.