In July, my Cato colleague Ari Blask and I wrote a study critiquing the National Flood Insurance Program.

We made what—to us—seem like obvious critiques of a broken program: it doesn’t charge an actuarially fair price for many homeowners who live in flood plains, and those who get the best deal seem to be the wealthy. It also fails to use updated maps that detail the current geography and risks to homeowners, and generally doesn’t charge enough to cover the costs of major catastrophes. The result of this is that we have too much development in the flood-prone areas of the country.

After the report came out I appeared on a few radio and TV shows and received a few emails about our research, and the main complaint I received—in fact, virtually the only feedback I got—was from people who said they were in a 100-year floodplain but had never seen any flood. They were angry that their bank made them buy this insurance to get a mortgage, and didn’t think it was fair.

On Thursday, I was a guest on another radio show to talk about the after-effects of Hurricane Harvey, and the second caller complained about… having been forced to purchase flood insurance despite his belief that his house couldn’t flood. The next caller chimed in with the same complaint.

Apparently not too many people see Harvey as being a precautionary tale. But that’s only natural: human beings aren’t great at perceiving risk. Surveys show that people worry much more about dying from a terrorist attack or a plane crash when slipping in the shower or getting hit by lightning are far more deadly, for instance.

However, insurance companies tend to be pretty good at discerning risk—they go bankrupt if they’re bad at it. Life insurers hire teams of people to try to understand longevity risk, for instance, and property and casualty insurers do the same to understand the risk to homes.

They use this information in pricing the cost of life insurance and homeowners insurance, and they have every incentive to get it right—if they charge too high a price they won’t get much business; too low and they will lose money.

The federal government—which administers the National Flood Insurance Program—does not have such a keen incentive to get prices right. In fact, in many regions of the country they use flood insurance maps that are decades out of date because it would anger homeowners—who are also voters—to force them to pay more than they think is fair.

I am not sure this is going to change. One thing that Harvey will do is bring more pressure on the federal government to provide “affordable” flood insurance to more people.

However, not all places merit cheap flood insurance—certainly not people who live along the gulf coast or along the North Carolina beaches, where flooding is not uncommon. These people tend to be well off already—many of these are vacation homes.

Harvey is unusual in that it created a monumental flood in a place where most of the homes were not in a flood zone. Some meteorologists described it as a “1,000 year flood.” It is not at all clear how the government will react to this disaster—hopefully it decides that it should be a priority to accurately measure flood risk, for starters. While the government is clearly not beyond subsidizing rich people (that’s what the mortgage interest deduction is all about, after all) perhaps we shouldn’t use this disaster to expand the National Flood Insurance Program and its inherent subsidies, as I fear people will begin to suggest.

Most people agree that we want a federal government to assist the public when we have unforeseen natural disasters, such as what has occurred with Hurricane Harvey.

However, the government’s role in the flood insurance market has exacerbated the damage done by catastrophic floods. By not charging the proper premiums for flood insurance, homeowners don’t make all cost-effective mitigation efforts and we see more development in flood-prone areas than would otherwise be the case.

As Congress begins to deliberate on the re-authorization of the National Flood Insurance Program, it is worth asking why the federal government feels obligated to provide flood insurance in place of the private market, and whether concentrating its efforts in disaster mitigation and relief, rather than poorly administering an insurance program, might be a more appropriate approach.

Having such a debate in the aftermath of a tragedy like Hurricane Harvey may lead the government to simply throw money at the problem. But the reality is that extricating itself from the flood insurance market would be the best thing the federal government could do in the long run if it wants to mitigate the damage caused by future floods, as well as its own obligations during such disasters.