We should not let the mild experience of the 2006 hurricane season dim our memory of Katrina, which killed more than 1,000 people, left hundreds of thousands homeless and caused more than $100 billion in economic damage.

Unfortunately, it is likely over the next 10 years that our country will face another catastrophe with at least as much devastation, and we have not made sufficient progress in managing catastrophe risk.

Significant segments of the infrastructure and property base in the United States are highly vulnerable to catastrophe losses. Our population centers have grown and likely will continue to grow rapidly in the areas most vulnerable to hurricanes or earthquakes: the Atlantic Coast and California.

Vulnerability to catastrophes

In terms of economic damage, Hurricane Katrina was the largest catastrophe loss on record in the United States. Even so, it was far from a worst-case scenario. Examples of other possible catastrophes with an even larger economic impact include events such as a Category 4 or 5 hurricane traveling through Galveston Bay to Houston, a major storm making landfall near New York or a major earthquake in California.

With the stakes so high, why has so little been done to prepare for natural catastrophes?

The answer seems to arise from the difficulty in understanding the risk profile. For example, AIR Worldwide Corp. estimates the probability of a Katrina-sized or greater economic loss affecting a small region of the Gulf Coast is less than 1% annually, or once in 100 years on average.

However, when calculating the probability that a hurricane would cause a loss of this size anywhere along the coast, the annual probability jumps to around 4%, or once every 25 years on average.

Finally, if we include catastrophe losses due to earthquakes, the probability that the United States will experience losses equal to or greater than those caused by Katrina is almost 10%, or once every 10 years on average.

So, there is a high probability that we will see another catastrophe at least the size of Katrina somewhere in this country over the next 10 years. It is tempting to take comfort from the low probability of a catastrophic event in a given region next year, but as a nation, we need to plan around the high probability of a major event occurring somewhere in the next decade.

Lessons to be learned

There are three fundamental and related policy questions that still need to be addressed:

c What limitations should our society set on growth in catastrophe-prone areas?

Because virtually all regions have some potential catastrophe exposure, it is not possible to limit growth to "risk-free" zones. However, it is clear that more care is needed: Even now, it is possible to build in very vulnerable areas, such as barrier islands along the coast, with inconsistent and often limited building standards.

The historical approach has been to rely on local governments for building codes. However, given the complexity of analyzing and managing the risk, and assuming that the federal government likely will bear a large financial burden to aid in the recovery from a major catastrophe, we advocate a larger federal role that includes minimum standards for building codes in catastrophe-prone areas.

c Who finances reconstruction after a catastrophe?

In general, given our capitalist economy, individuals and businesses should bear responsibility for financing the reconstruction of their properties. Assuming an adequate appreciation of the risk, and a free market pricing mechanism for insurance that does not exist in many states, then homeowners and businesses should see substantial cost to locating in catastrophe-prone areas, and will only do so when they judge the benefits to outweigh the cost. Individuals must recognize the scope of the risk and know that they will bear the responsibility for being uninsured or underinsured. (Note: In California, fewer than 15% of homeowners currently buy earthquake insurance.) Lower-income homeowners may not have the means to assume the risk, and some governmental support may be needed.

Some argue for a federal backstop for insurance companies in the event of a large catastrophe, but the absence of systemic failure from Katrina suggests that such a mechanism may not be needed, or suggests that the magnitude of the triggering loss would need to be larger than the loss from Katrina. We should avoid or at least minimize any mechanism that would subsidize or otherwise encourage settlement in catastrophe-prone areas.

c How can we create the most effective role for government?

Even with a bias toward private-sector solutions, insurance likely will pay for less than 50% of the total costs after a megacatastrophe. There will be a large role and cost for government arising from the immediate disaster response, to help lower income citizens and to repair uninsured government buildings and infrastructure. Mechanisms for dealing with catastrophes need to be created now and the question of the relative roles of federal and state government in disaster relief must be clarified.

Strategy needed

Some states with significant experience in dealing with catastrophes have developed robust systems for managing catastrophe losses and disaster recovery. However, the magnitude of catastrophes and their relative infrequency within other geographic areas indicates the need for a larger federal role in developing and guiding a comprehensive strategy for mitigation and disaster recovery.

Left unchecked, property value growth in vulnerable areas will result in a doubling of the catastrophe loss potential over the next 10 years. It is time that we address the fundamental policy questions of catastrophe risk, or we will repeat the mistakes of the past.