Hsiang and Jina looked at 6,712 cyclones, typhoons, and hurricanes observed from 1950 to 2008 and the economic fortunes of the countries they struck in the years that followed. With their data, Jina and Hsiang can decisively say: These storms are bad—very bad—for economic growth.

"There is no creative destruction," Jina told me. "These disasters hit us and [their effects] sit around for a couple of decades." He added, "Just demonstrating that that was true was probably the most interesting aspect for me to start with."

Hsiang and Jina find that such storms (which they group under the umbrella term "cyclones") can be as bad as some of the worst sorts of man-made economic challenges. A cyclone of a magnitude that a country would expect to see once every few years can slow down an economy on par with "a tax increase equal to one percent of GDP, a currency crisis, or a political crisis in which executive constraints are weakened." For a really bad storm (a magnitude you'd expect to see around the world only once every 10 years), the damage will be similar "to losses from a banking crisis." The very worst storms—the top percentile—"have losses that are larger and endure longer than any of those previously studied shocks."

Here's a little chart they made comparing these different sorts of disasters:

(Hsiang and Jina)

The effects are lasting: Overall, they find that "each additional meter per second of annual nationally averaged wind exposure lowers per capita economic output 0.37 percent 20 years later" (emphasis added). Put simply, economies "do not recover in the long run." 

Put simply, no civilization should locate in hurricane territory.