Many of the most expensive flood and storm disasters in US history have occurred in recent decades. The glib response is to blame the severity of these catastrophes on climate change, but are we looking in the wrong direction?
Almost 20 years ago, the National Wildlife Federation issued a report on this subject, Higher Ground. It argued, among other things, that federal flood insurance was amplifying the impact of storms by encouraging Americans to build and rebuild in areas prone to flooding.

Higher Ground revealed a home in Houston that was valued at $114,480, but which had been flooded 16 times and thus received $806,591 in federally subsidized flood insurance payments.
Michael Grunwald, writing in Politico, sums up the issue:

Houston’s problem was runaway development in flood-prone areas, accelerated by heavily subsidized federal flood insurance.… Congress often discusses fixing flood insurance to stop encouraging Americans to build in harm’s way, but the National Flood Insurance Program is still almost as dysfunctional as it was 19 years ago. It is now nearly $25 billion in the red, piling debt onto the national credit card. Meanwhile, cities like Houston — as well as New Orleans, which Higher Ground identified as the national leader in repetitive losses eight years before Hurricane Katrina — continue to sprawl into their vulnerable floodplains, aided by the availability of inexpensive federally supported insurance.

Those who advocate for insurance subsidies might say that every great city was once small and population growth requires building. The same subsidies advocate might say that to take a position against subsidies is to prevent growth and push people into rural areas where they prefer not to live.
But the case against subsidized flood insurance is not a case against growth; it is a case against distorted growth. Federally supported insurance overrides the risk-reducing incentives that insurance premiums provide and results in building in vulnerable areas. For example, wetlands are lost to distorted growth.  From 1992 to 2010, Northwest Houston lost more than 70% of its wetlands.
Grunwald is clear that it is subsidized flood insurance, not climate change, that is responsible for Houston’s disaster:

Two relatively modest storms that hit Houston in 2015 and 2016 — so small they didn’t get names — did so much property damage they made the list of the 15 highest-priced floods in US history. But Houston’s low-lying flatlands keep booming.

Given the subsidized flood insurance, should we be surprised that millions of Americans live in harm’s way, on flood plains? In short, as Grunwald observed, “Floods are not really natural disasters. They are natural events, but they are mostly man-made disasters.”

How insurance works in a free market

Every day we make decisions based on risk. You may not know it, but insurance premiums provide incentives to reduce risk. In a free market, insurance premiums on cars, for instance, tend to settle toward an “actuarially fair price.” This is based on the insurance companies’ estimates of the chance you’ll make an insurance claim on your car, as well as the possible loss to the company if you do.
If you have a history of drunk driving, that increases the chances you’ll make an insurance claim on your car — so your premiums will be higher, and that encourages you not to drive in the future (or to drive sober in the first place).
If you bought a fancy sports car, that would increase the loss to the company if your vehicle gets damaged — so your premiums would be higher. That might encourage you to buy a less-expensive vehicle.
If governments subsidized car-insurance policies the way they do with flood insurance, we’d see more drunks driving Ferraris around, at the expense of taxpayers.

If the risk is high, premiums will be prohibitive.

Getting the government out of the flood insurance business and having insurance companies determine actuarially sound premiums is the only way for homeowners, businesses, and builders to know the real risk they are assuming. If the risk is high, premiums will be prohibitive, and building will migrate to areas less prone to flooding.
Unfortunately, the prospects for reform are low. Eliminating the National Flood Insurance Program is impossible politically. How likely are even modest policy reforms about premiums or redrawing flood plain maps to better reflect real risk? Such proposals are drowned out by politicians who score easy points by blaming the damage on climate change and passing popular rebuilding bills.
The sad, but most likely, prospect for the near future is that taxpayers will spend hundreds of billions rebuilding Houston in areas prone to suffer flood and storm damage again and again.