Insurance Against Terrorism

An alternative to unlimited liability for taxpayers

by Eli Lehrer
08/10/2009, Volume 014, Issue 44

After hijackers destroyed the World Trade Center on 9/11, taxpayers ended up spending a lot of money to aid the injured, rebuild public infrastructure, improve security, and help the jobless. But the private firms with property and workers in lower Manhattan fell back on their private insurers. And the companies paid out: Over $35 billion flowed from their capital reserves to people harmed in the attack. No insurers went under as a result of 9/11 and all but a handful of claims were paid within a few months. In short, it was a shining hour for the insurance industry.

But if another major terrorist attack takes place, the industry won’t have as much need to step up to the plate. Instead, the government will take charge. Under an obscure but potentially budget-busting program-terrorism risk insurance-the federal government has assumed nearly unlimited liability for major terrorism losses. The program, called TRIA, can claim broad support but has deep flaws and imposes billions in liabilities on taxpayers. The program, though never intended to be permanent, will be entrenched before long. Still, it’s not too late to restore an affordable, private system of insurance.

Under the current program, once industry-wide private commercial and workers compensation insurance claims from a terrorist attack exceed $27.5 billion, TRIA kicks in and covers the remaining expenses up to $100 billion. (Congress would almost certainly lift the $100 billion cap if claims exceeded that amount.) In theory, the money to pay claims would come from a tax (up to 3 percent) on just about every eligible insurance policy in the country. If this tax proved insufficient, Congress would have to use other revenues.

Although TRIA was created in 2002 as a post-9/11 stopgap, insurance companies have shown almost no interest in replacing it. Often fractious industry groups representing brokers, insurers, reinsurers, and commercial insurance consumers have lined up in support of the program. And, when the Government Accountability Office studied terrorism insurance earlier this year, it found that the chances of a private terrorism insurance market developing were very slight. TRIA is currently authorized through 2014.

And that’s a problem because the federal government-already stretched with bailouts and “stimulus” spending-has no business running a hugely expensive insurance program. Its record isn’t encouraging. The other major federal effort at disaster insurance, the National Flood Insurance program, owes the Treasury about $19 billion, has no way to pay it back, and has actually increased the nation’s susceptibility to flood by effectively subsidizing building in flood-prone areas. States like Florida that attempt to run property insurance programs have done even worse.

But that doesn’t mean that doing away with federal terrorism insurance will be easy. The insurance industry has a good reason to support it. The current system for writing insurance really can’t deal with terrorism adequately.

Explaining why this is so requires some background on how insurers manage risk. To write a policy, an insurer will build a group of similar risks-a pool-unlikely to experience losses at exactly the same time. An insurance company might calculate that the chances of a $100,000 house burning down during a given year were 1 in 100. It could then write policies for 100 homes in different neighborhoods worth $100,000 each and charge a yearly premium of $1,200 for each policy. Of the $1,200 collected, $1,000 would cover expected claims and the extra $200 would cover the expenses of writing the policy, provide for the purchase of reinsurance (insurance for insurance companies), build reserves, provide return on capital for company owners, and offer a margin of safety for the insurers’ own uncertainty about its “1 in 100 chance” calculation.

But the actuaries who do these calculations can’t make decent guesses about the likelihood of terrorist attacks. The past two decades have seen only three significant terrorist attacks on American soil. For every obvious target (like the World Trade Center), terrorists have picked a less-obvious one (such as Oklahoma City’s Murrah federal building). The best information about terrorist risks, furthermore, remains a closely guarded secret within the intelligence and law enforcement communities. Before 9/11, large commercial insurers and reinsurers generally provided terrorism coverage nonetheless.

It’s easy to see why they stopped. Based on existing data on the number of modern attacks on office buildings and the number of office buildings in the United States, actuaries forced to make the calculation might guess that the average yearly chance of a terrorist attack on a $25 million office tower located in a midsized city, is one in a million. This would indicate an “actuarial” yearly premium in the neighborhood of $25 for $25 million of coverage. But no insurance company with competent management would ever risk $25 million in capital for a premium of $25 or even $2,500. Even if an insurer decided to sell terrorism insurance at a price like that, state insurance regulators would likely block the sale as too risky to the company’s other business. Insurance priced high enough to satisfy regulators and insurance company managers, on the other hand, probably wouldn’t find any buyers.

Clive Tobin, the CEO of the Bermuda/London reinsurer Torus and a longtime reinsurance executive has floated another plan in comments at trade conferences: true reciprocity. Under a “true reciprocal” system, 25 firms that each own a $25 million office building would each take responsibility for paying $1 million if terrorists destroyed any group member’s building. The firms would pay no yearly premiums in return for the coverage (they might pay administration fees and exchange $1 payments to make the contracts between themselves legal) and would not be regulated as insurers. Instead, they would simply pledge their full faith and credit to pay the claim if another group member experienced an attack. “What you are really looking for is an agreement,” says Tobin, “to avoid an insured having to tell their board that their location has just been destroyed and they have no insurance.”

The idea, as Tobin conceives it, could have some other wrinkles. For example, a company that owns a building in Manhattan might take on $50 million of risk for a Minneapolis-based company in return for the Minneapolis company taking on $10 million in Manhattan risk. Some sort of formal exchange, very likely, would have to exist to match participants’ risks.

However it works out in practice, the idea has enormous potential benefits. Neither insurers nor insurance purchasers would have to divert any capital to buy expensive insurance policies against the unlikely possibility of terrorist attack but, simply by expanding the size of the groups they joined, could reduce their liabilities. Taxpayers would owe nothing. (Tobin suggests a secondary backstop that would have the government provide partial coverage against a specific company’s default on its reciprocal obligations.)

And the idea isn’t new. In fact, many existing mutual insurance companies like Ohio’s Westfield Group and San Antonio-based USAA started writing farm and automobile insurance exactly this way and sometimes retain a few legal structures of reciprocity. The idea faded from practice as an insurance company that charges premiums in advance has a much easier time making payments on claims and can generate more profits by investing premium dollars between claims. If it doesn’t suit most types of modern insurance, however, such a structure seems almost ideal for an entity providing sizeable commercial enterprises with coverage against rare terrorist attacks.

Current law, however, makes it almost impossible to set up such a structure. “The major [problem] for me is how to make sure this process can be executed in an appropriate legal and regulatory framework, but keep this at a light touch,” Tobin told me. Since each state regulates insurance individually, at least some would likely demand that every party participating in a reciprocal system submit to regulation as an insurance company. And noninsurance firms would never agree to that.

Risk retention groups, a class of lightly regulated federally authorized insurance companies that focus on malpractice coverage and backing for consumer product repair insurance, have something in common with Tobin’s idea but do typically charge annual premiums and face all sorts of restrictions likely inappropriate to Tobin’s idea. Just as important, the apparently permanent existence of TRIA and the lack of a real potential for profit in its absence makes it unattractive for anybody (even Tobin’s own company) to spend lots of resources pushing an alternative.

Making an alternative to TRIA work will probably require special legislation in Congress. Given the uncertainties, furthermore, it would be unwise to repeal TRIA before reciprocal terrorism insurance arrangements get off the ground. Piloting the idea alongside TRIA, particularly by starting in areas unlikely to experience terrorist attack, could provide an important proof of concept.

The idea needs further refinement. But it’s well worth considering. Replacing TRIA with a private system won’t be easy. But leaving TRIA in place sticks the American taxpayer with nearly limitless liability for the coverage of private property. If the government would only facilitate its emergence, however, it seems that a private solution could be found to pay the bills for terrorist attacks.

Eli Lehrer is a senior fellow at the Competitive Enterprise Institute, where he directs the Center for Risk, Regulation, and Markets.