MARIO TAMA/Getty Images News     

   Photo: The French Quarter on August 29, 2005

Many people are probably wondering today why, two years after Katrina, New Orleans remains something a little less than a shining city on a hill. The news on the Big Easy's recovery is not all bad, but it's certainly disappointing for those of us who were hoping the city would bounce back quickly from tragedy. Only the old parts of New Orleans, which were built on the higher ground and were never destroyed, seem to be thriving—and many people have fled for the suburbs. In a fascinating New York Times Sunday Magazine article about the wild world of catastrophic insurance, Michael Lewis goes a long way toward explaining what is going wrong:

Louisiana cannot generate and preserve wealth without insurance, and it cannot obtain insurance except at the market price. But that price remains a mystery. Billions of dollars in insurance settlements — received by local businesses and homeowners as payouts on their pre-Katrina policies — bloat New Orleans banks and brokerage houses. The money isn't moving because the people are paralyzed. It's as if they have been forced to shoot craps without knowing the odds. Businesses are finding it harder than ever to buy insurance, and homeowners are getting letters from Allstate, State Farm and the others telling them that their long relationship must now come to an end. "I've been in the business 45 years," says a New Orleans insurance broker named Happy Crusel, "and I've never seen anything remotely like this." An entire city is now being reshaped by an invisible force: the price of catastrophic risk. But it's the wrong price.

Lewis's article is basically a long profile of John Seo, a math whiz who has pioneered "esoteric financial options"—complex financial products that other people couldn't figure out how to price properly. Seo's insights on how to spread the risk from catastrophic events such as Katrina are hugely important in an age of worsening storms. But they could have unexpected pernicious consequences.

Consider the undiminished risk of flooding in New Orleans, wildfires in Malibu, or hurricanes along the Florida coast. I know; everybody wants to be near water. But the truth is, people shouldn't be building their homes in flood plains, in areas that are especially prone to severe wildfires and mudslides, or on ecologically fragile barrier islands—and insurers shouldn't be encouraged to sell policies to people building new homes in such places. (I'm not calling for, say, the wholesale evacuation of Singapore.) It is politically costly for politicians to resist massive bailouts after events like Hurricane Andrew; just look at what happened to Bush I in the 1992 election. Knowing this reality, insurance companies might take risks that they otherwise wouldn't. We need to ensure that Seo's innovations, for all the good they might do, don't magnify this problem.