#8 Why the Hurricane Plan Got Trashed

Wired 13.11: START

"Everybody talks about the weather," Mark Twain famously noted, "but nobody does anything about it." Sounds a lot like Hurricane Katrina. The experts in New Orleans and beyond talked about the potential for a levee collapse as far back as the 1960s. A 2001 study by the Federal Emergency Management Agency predicted most of what actually came to pass. And last year, National Geographic published a stunningly prescient step-by-step description of what ultimately befell the Crescent City. By the time Katrina roared ashore, every detail of the catastrophe had been foretold.

But the test of good scenario planning is not just better predictions but better decisions. And the decisionmakers at every level failed to make the right calls in Louisiana. As a result, confusion reigned, more than 1,000 people died, and half a million homes were damaged or destroyed. The disaster will be etched in the nation's memory along with the 1871 Chicago fire, the 1889 Johnstown flood, and the 1906 San Francisco earthquake.

Turns out there's a huge difference between anticipating disaster and actually being prepared for it. Scenario planning is not a waste; computer models can now game the behavior of millions of variables and render nuanced predictions of everything from bioterror attacks to massive earthquakes. Too bad that when it came to a hurricane like Katrina, political expediency and shortsightedness prevented adequate preparation. Everyone knew what was coming; they just failed to act on it.

Why? Cost, for starters. Planning for the worst is expensive, financially and politically. Government officials have a clear interest in pleasing constituents with projects and benefits that make a difference within one election cycle. Disaster prep gets pushed down the priority list. This year, for example, Congress and the Bush administration cut funding to improve levees in New Orleans from $27.1 million - the amount requested by the Army Corps of Engineers - to $5.7 million. The result: Construction was postponed indefinitely.

Even when politicians are willing to take the threat of disaster seriously, the perverse incentives of short-term thinking can warp the effectiveness of the spending. In the immediate aftermath of 9/11, Congress allocated $10.6 billion to homeland security. But instead of doling it out to the places most likely to be attacked, the money was widely distributed - effectively spreading the pork among congressional districts. Ranchers in Wyoming are getting more protection per capita than residents of targets like New York City and Washington, DC.

Then there's the insurance mentality. The brutal logic of risk dictates that it's cheaper to pay for the aftermath of a single, relatively localized disaster than it is to prepare for that same disaster in every place at every time. This is how insurance companies make money. But unlike government, they have actuaries who are able to accurately estimate value and thus assess risk. If the true value of New Orleans - especially its importance as a port - had been understood by government officials, the cost to prepare properly for a hurricane the strength of Katrina would have seemed a bargain.

Even when governments do face up to the likelihood of extreme disasters, politics can still undermine the results. In 1979, California Governor Jerry Brown, a Democrat, established an earthquake-preparedness task force that was a model of prudence and foresight. (Disclosure: I worked for Brown setting up the panel.) After Republican George Deukmejian was elected in 1982, he reversed many of the measures Brown had put in place. In 1985, he disbanded the task force. The people of the Bay Area paid for that lack of preparation in 1989 when the Loma Prieta earthquake caused a freeway collapse, fires in San Fran?cisco's Marina district, and damage to the Bay Bridge.

All of this does not mean that scenario planning can't work or that attempting it is useless. As Jeanne, a Category 3 hurricane, approached the Florida coast in September 2004, computer modeling experts from Los Alamos National Laboratory supplied state and federal emergency teams with detailed - and uncannily accurate - predictions on power outages, flooding, and other storm effects. The result: Mandatory evacuations moved people safely out of what became the hardest-hit areas, and repair crews descended on damaged electrical centers almost before they went down. Power was restored much more quickly than it otherwise would have been. It's a modest example, but considering how wrong things can go in the face of a hurricane, it's a scenario worth following.

Peter Schwartz (peter_schwartz@gbn.com) chairs the Global Business Network and helped create the field of scenario planning.