Weather and Society Watch
An Economic Postcard from Sunny (So Far) Miami: June 1, 2010
Case in point: Today marks the first day of the 2010 hurricane season. While 2009 was quiet, with no U.S. landfalling storms, the National Weather Service forecasts an active 2010 season. Already the first tropical storm of the season, Agatha, has killed at least 180 people in Guatemala , Honduras and El Salvador over the Memorial Day weekend . Earthquake-ravaged Haiti and the Deepwater Horizon oil spill are particularly vulnerable to extreme weather. Are we ready for whatever nature may have in store?
In many circumstances, the economic contribution of weather, good or bad, may be subtle; in a few others, though, weather may have devastating societal impacts. Either way, be it a thousand rained-out fishing trips or a single tornado, there's no escaping the economic importance of weather. While the economic value of most weather-related experiences may be small, their sum total surely is not, judging from the estimated 300 billion weather forecast accesses in the United States each year, which implies a value of current weather forecast information of $286 per U.S. household per year, or $31.5 billion in total value to U.S. households (Lazo et al. 2009). Weather affects crop yields, recreational decisions, construction activity, energy use, transportation, and so on. In particular, how we manage the economic risks posed by extreme weather is crucial for many regions, including the southeastern United States where I live.
Preparation is the essence of how to lessen the likelihood and severity of severe weather, such as hurricanes. We may prepare for hurricanes in a variety of ways, including our choices about where we live, what sort of structures to build, and how to insure them. Each of these economic decisions poses important challenges, but for brevity and because of its linkages with the ongoing global financial crisis, I will focus here on property insurance.
As recent events have shown, modern economies cannot function without insurance, which disperses risk, making loss—when it occurs—manageable. Insurance payouts also provide recovery resources in the aftermath of a disaster. Often insurance is required for investments in physical property, since it facilitates the use of the property as collateral for loans. In short, an insurance crisis would be nothing less than a full-blown economic crisis. Unfortunately, that's where we seem to be going in the southeastern United States . The rapid escalation of hurricane insurance claims over recent decades is nothing short of alarming; since 1985, insurers have experienced negative cumulative profits for coastal properties extending from North Carolina to Texas (Michel-Kerjan 2009). Insured catastrophic losses are doubling roughly every 10 years (Pielke et al. 2008).
Not surprisingly, the root problem with hurricane insurance is an inadequate spread of risk. Insurance is predicated on pooling; it works effectively when losses from any individual event are embedded in a pool of funds so large that the loss from any single event is manageable. Yet that is far less likely for events such as hurricanes that affect large numbers of policy holders at once. What is surprising is the sizable, additional capital needed for insurers of events, such as hurricanes, which pose correlated risks.
The hurricane insurance problem and the correlated risk that underlies it are perhaps best understood as an instance of Pareto's Law, more commonly known as the "80-20 rule". In 1906, the economist Vilfredo Pareto discovered that 20% of the Italian population owned 80% of the wealth. Stated more generally, it suggests that , for many events, roughly 80% of the effects come from 20% of the causes. The idea is that extreme and rare events, in complex systems such as economics and meteorology, have a far greater than expected impact. For anyone used to thinking of averages as being most useful in characterizing possible outcomes (i.e., bell curves), Pareto's Law is a radical departure. Yet the 80/20 rule does appear to fit common sense better than what we were told in Statistics 101. Most of the total movement in any stock over a single year is often attributable to abrupt changes on a few select days. Similarly, in any decade, a handful of the strongest hurricanes do more property damage than the rest put together.
Natural and economic hazards have eerie similarities. The worst event is twice as bad as the second, which is double the next, and so on. In response, we should train ourselves to expect hazards to arrive not gradually but in sudden discontinuities that reshape us. Like the hurricane seasons of 2004-05, the ongoing global financial crisis serves as a reminder that we ignore our exposure to systemic, correlated risks at our considerable peril. (Remember hearing, "Housing markets can't all tumble at the same time"?) Those who have lacked the discipline to avoid risks that they did not understand have destroyed wealth on an unprecedented scale. Economic history suggests that it will happen again, unfortunately. One catastrophe modeling firm has estimated that a large hurricane in southeast Florida could cause insured losses of $130 billion and a total economic loss of $260 billion (USGAO 2007). 
A numerical example helps show just how big a problem correlated risk is for insurers. Consider Rade Musulin's comparison of two insurance companies, one writing hurricane coverage (correlated risks) and the other writing for fire (independent risks). Suppose that the hurricane insurer can expect a single $100 million loss per 100 years, while annual fire losses will vary between $0.8 million and $1.2 million. Annual premiums of $1 million would cover each sort of expected loss. But to meet claims in any given year, the fire insurer would need $200,000 in capital. The hurricane insurer, however, would need $99 million for the single $100 million loss that occurs each 100 years (Musulin 1997).
Catastrophic insurers face spatially correlated risks, which require significantly more capital for the insurer to protect itself against large losses. Because hurricane damages tend to follow Pareto's Law, hurricane insurers must be prepared to pay out in any given year for the 20% of the storms that may cause 80% of the damages. Those in Florida and elsewhere attempting to limit the cost of windstorm coverage confuse it with fire insurance and, in doing so, mistake a risk subsidy problem for what is actually a risk pooling problem. Too cheap insurance, in turn, only makes matters worse, by encouraging further coastal migration.
Can our insurers (and those who regulate them) deliver? As I said, I'm nervous this hurricane season. Future hurricanes will come, and we must decide in advance how we will pay for them. The crucial question is whether we wish to pay for hurricane damages ourselves in advance or whether to shift that burden to someone else. While my windstorm coverage is still affordable, that means little if my insurer (the State of Florida , in my case) becomes insolvent in a crisis.  Insurers sell a promise, one crucial to the flow of credit, whose value comes from insurers' ability to pay claims. While the sun is out today, Florida may be one major hurricane away from needing a Federal bailout. Failure to appreciate how weather-dependent we are in our economic lives can be expensive, indeed.
*David Letson (firstname.lastname@example.org) is a professor at the University of Miami 's Rosenstiel School of Marine and Atmospheric Science (RSMAS).
Lazo, J.; R. Morss, and J. Demuth. 2009 . "300 Billion Served: Sources, Perceptions, Uses, and Values of Weather Forecasts" BAMS 90 : 785-798.
R.T. Musulin, 1997. "Can Our Children Afford 'Affordable' Insurance?" Emphasis 2 : 2-5.
Pielke, R.A. Jr.; J. Gratz; C.W. Landsea; D. Collins; M.A. Saunders; and R.T. Musulin 2008. "Normalized Hurricane Damage in the United States : 1900-2005" Natural Hazards Review 9(1) : 29-42.
U.S. General Accountability Office. 2007. "Natural Disasters: Public Policy Options for Changing the Federal Role in Natural Catastrophe Insurance." Report to the Ranking Member, Committee on Financial Services, House of Representatives. www.gao.gov/new.items/d087.pdf . Accessed: 2 April 2010.
1 One rule of thumb is that insured losses, which are carefully measured, tend to be about half of overall economic losses, which are estimated with great difficulty.
2 The Florida Hurricane Catastrophe Fund, the State-run re-insurer, faced an $18 billion shortfall as a worst-case scenario for the 2009 season and faces a $7 billion worst-case shortfall for 2010.