Evacuations and Uncertainty
Irma spent the night slow dancing with Cuba. Very bad for Cuba but probably spared Miami and the East coast of Florida the worst effects of the hurricane. If so, those who evacuated Miami at considerable cost and trouble may be outraged. The thing always is that prediction of hurricane path and especially intensity, though drastically improved, is not, and is not likely to become, an exact science. On the other hand, if evacuations had not happened, and the quite likely event of a direct hit on the East coast had happened, the casualties could have been immense. All of which invites the question: is there a better way?
I think there is. It's not cheap, but I think it would save lives and money over the long run. I've mentioned the basic idea before. Build large, well-equipped, durable, and multi-use shelters near as many flood prone regions of high density population as possible. This should be accompanied with two other policy changes: phase out flood insurance and discourage building in flood prone regions.
It's simply not feasible to evacuate millions or tens of millions of people for hundreds of miles. Such evacuations are dangerous, expensive, and frequently leave the evacuees worse off, for example, consider the East Floridians who evacuated to West Florida. If safe, well-equipped shelters were available within a few miles, people could much more safely, easily and quickly evacuate. Moreover, it would be both safe and sensible to wait until forecasts were far more certain.