Scott Sumner  

The root causes of natural disasters

Larry White on the Origin of M... Tristan Caplan's Tetlockian Gl...

When an area is hit by a natural disaster, the media tends to focus on the proximate cause---a storm, earthquake or some other unusual event. But in many cases there are also deeper causes, which offer a more powerful explanation. Thus the proximate cause of a North Korean famine might be a drought, whereas the deeper cause is bad economic policies. The proximate cause of the high death toll in an Italian earthquake is the collapse of buildings, whereas the deeper cause might be something like corruption in the building industry. For the recent financial crisis, the proximate cause was a lot of loan defaults, whereas the deeper cause was moral hazard created by FDIC, Too-Big-To-Fail, and the GSEs.

The Houston area is currently suffering from a very costly disaster:

Flood damage in Texas from Hurricane Harvey may equal that from 2005's Hurricane Katrina, the costliest natural disaster in U.S. history, said an insurance research group on Sunday. . . .

Hurricane Katrina resulted in more than $15 billion in flood insurance losses in Louisiana and Mississippi that were paid by the National Flood Insurance Program (NFIP), a federal program that is the only source of flood insurance for most Americans.

But Andrew Siffert, the resident Meteorologist at insurance broker BMS Group said it was easy to understand that Harvey will easily reach well over $10 billion in economic loss, and estimated the insured loss at over $5 billion. That figure did not include money that would be paid by NFIP, he added.

The NFIP is already deeply in debt and likely will have to be bailed out again by U.S. taxpayers, as it was after Katrina, to cover the bill for flood damage claims from Harvey.

Just like FDIC, the NFIP program creates excessive risk taking. It also leads to the development of fragile coastal areas, which is why it is opposed by an unusual coalition of free market advocates and environmentalists:

The NFIP was created in 1968 after private insurers stopped selling flood coverage. Critics have said the program provides a misguided tax subsidy to coastal and river valley property owners, encouraging development in flood-prone, often environmentally sensitive areas such as wetlands.

Congressional reform efforts, supported by a coalition of environmental activists and free-market advocates, have largely been thwarted by waterfront real estate interests.

In this particular case, I doubt whether the NFIP played a major role in the disaster, as much of the damage has occurred far from the ocean. But in the case of other coastal storms, the role of flood insurance is much more important, as it encourages the construction of vacation homes on outer islands that are easily overrun by hurricane level storms.

There's another more subtle problem with the Federal government providing disaster relief. This takes pressure off of states and local governments to take preventive actions. If Louisiana had known that it would have had to foot the entire $15 billion tab for Hurricane Katrina, then they might have done more preventive work such as improving the levies along the river.

In the case of Houston, it's too late to fix the federal insurance policy. Going forward, a good first step might be to turn the insurance function over to the states, perhaps accompanied by some sort of block grant. States would then have the incentive to explore options that would minimize the risk of catastrophic insurance losses.

PS. Good luck to everyone on the Texas coast--I have some relatives down there.

Screen Shot 2017-08-28 at 2.27.47 PM.png

Comments and Sharing


COMMENTS (16 to date)
Andrew_FL writes:
Going forward, a good first step might be to turn the insurance function over to the states, perhaps accompanied by some sort of block grant. States would then have the incentive to explore options that would minimize the risk of catastrophic insurance losses.

Judging from our State property insurer here in Florida, probably not.

Rebes writes:

Doesn't the root cause go further back in history than national flood insurance?
Many of our largest cities grew in very challenging, if not irrational, locations. San Francisco and Los Angeles are on a crack, Phoenix is in the desert, Houston is in a hurricane path and in a swamp, New Orleans is below water level, Miami is in a hurricane path, Washington is in a swamp - and on the list goes. Why is that? Did it conform to the pioneer attitude of the country to build cities in challenging places?

Joe writes:

I'd prefer to take it a bit further and let the market handle insurance for floods. Let private insurers do this. If a private insurer won't insure a property against flood, then it is cause to not build it in the first place.

Matthias Goergens writes:

Keep in mind that subsidised insurance coverage will be reflected in higher rents and land prices.

So these subsidies are really only a gift to whoever owns the land when they are first credibly announced (thanks to efficient markets we don't have to even wait for them to be implemented).

The current inhabitants don't benefit at all.

(There's a bigger principle at work here: it's very hard for the government to just give anything away as a simple gift.)

Alan Goldhammer writes:

I think your Katrina costs are wrong. The $15B was for the repair of the New Orleans area flood control system of levees, dikes and pumping stations. The over all cost to both LA and MS was $150M. The point you made about the states obligation to pay for the repairs is valid. I find it interesting that a lot of states argue for a diminished federal government role but when a disaster happens they want the assistance. Many of these states have very low tax rates and if they had to bank 'disaster relief' money that would have to change in a hurry!!

Haywood Jones writes:

Your comment about states and localities internalizing the costs of flood control structures, like levies, is wrong. (i.e. Louisiana footing the entire Katrina bill). Most major flood control structures are managed by the US Army Corp of Engineers and the projects are tendered by the Federal government.

Disaster relief on a state or local basis only is justifiable if the infrastructure funding is not channeled through the feds. Leaving aside the legal aspects, there are economies of scale in having FEMA - which can respond nationwide.

While Texas is huge and large portions are unaffected by hurricane Harvey and the associated flooding, how could a small state muster a rapid and thorough response if the entire state was inundated?

ZC writes:

"The current inhabitants don't benefit at all."

Do away with federally subsidized flood insurance and do you think the value of those properties will increase or decrease substantially? Yep, current owners benefit, because if the insurance goes away, many of them will probably end up 'under water'.

Alan Goldhammer writes:

Haywood Jones writes:

"Your comment about states and localities internalizing the costs of flood control structures, like levies, is wrong. (i.e. Louisiana footing the entire Katrina bill). Most major flood control structures are managed by the US Army Corp of Engineers and the projects are tendered by the Federal government."

There is a long history of mistaken management by the Army Corps of Engineers. Yes, these funds come from the Federal treasury, but why should that be the case? We know full well that many of the states in the southeast region get more money back from the US Government than they pay in taxes. Why should I as someone living in Maryland underwrite projects in states that don't adequately provide financing for them?

You cannot have it both ways.

Zeke5123 writes:


The fact that removal of the subsidy causes harm does not suggest the subsidy benefits the current resident. Consider the following:

P buys black acre for 70 dollars. Black acre has historically been subject to an easement. Absent the easement, blackacre would be worth 100 dollars. The legislature passes a law eliminating easements of this kind.

This results in a windfall of 30. I think you'd agree this was a windfall because P bought something worth 70 and obtained something worth 100. If you think the easement relief was a windfall, then it would be odd to say the easement was harming P.

In the same way, if P purchases whiteacre for 130 dollars and white acre benefited from an easement worth 30 dollars, you would t say P benefits from the easement; P paid for it. Removal of the easement would result in a windfall loss for P.

SCOTUS' takings jurisprudence has a name for this concept -- investment backed expectations. Now, the court butchered this concept in Penn Central. But as an analytical framework, it makes sense.

Haywood Jones writes:

Alan Goldhammer:

Don't mistake recognition of the facts for advocacy. The Corp of Engineers central role in infrastructure is one of many government bureaucracies run amok, mission creep at its finest.

I'd love for the Corp of Engineers to step out of most of its roles, although I doubt most states will do a better job. The Army tends to be one of the more competent government organizations.

Let's eliminate deductions for state and local taxes while we're at it. Why should I, as someone living in a state with no income taxes subsidize services for states that tax and spend excessively. Let's ensure no federal payouts for state & local public employee pensions too.

On flood insurance:
Many people purchased homes based on flood insurance rates at time of purchase. Mortgages require flood insurance. If subsidies are substantially reduced, the public outcry will be enormous. Lots of swing states are prone to flooding. Guess where most of the donors to both parties live - in nice houses on the water. The political calculus is very different than the economic one. Much as I would like the subsidies to stop, it's highly unlikely.

Floccina writes:

I think similarly about health insurance, the more the fed gov pays for health insurance the less the incentive for the state healthcare regulators to regulate with costs in mind.

High healthcare spending looks like a state problem to me, because per capita healthcare spending in lower in Utah than in Canada and just a little higher in Arizona than Canada. (Note: Utah is a little younger than Canada but Arizona is not.)

I think the feds should turn all the money that they now spend in healthcare to the states on a per capita age adjusted basis and tell the states you must cover the old people, poor people all federal government employees.

Seems bad to me to regulate at the state level and spend at the federal level. The feds comparative advantage is in collecting taxes, because it's easier to change states than countries.

Floccina writes:

Also I would settle for the Feds charging more for flood insurance. Enough to cover house replacement and the clean up after their fuel and stuff ends up in the water. And maybe even a little profit.

Thaomas writes:

A better understanding of this problem would require looking into why there is was no private insurance market before the Federal program.

This might involve the problem of self-selection, the property owner knowing her probability of being flooded better than the insurer. Or it could involve "too wet to fail," property owners knowing that -- whatever the shortcomings of public relief -- they will be partially "bailed out."

Scott Sumner writes:

Haywood, I don't agree. Even if the Feds built some of these structure, we would get a more efficient outcome if they were maintained locally.

There may be a few cases were FEMA is needed in disaster relief (although generally most of the relief comes from local sources) but surely the cost of the disaster should be borne as close to home as possible.

The Original CC writes:
The NFIP was created in 1968 after private insurers stopped selling flood coverage.
Scott, do you understand why this happened? Private insurers couldn't just raise their rates? I can't imagine that the asymmetric information problem is *that* bad. Or did people just assume that without insurance, they'd personally get bailed out by the gov't (which seems to often be the case)?

Like another commenter mentioned, it'd be nice if the federal gov't just charged proper prices for flood insurance.

TinMan writes:

When it comes to governments taking preventive action or buying insurance, there's a built-in disincentive for the players at the table: they'll incur cost and (most likely) get no direct benefit because they'll be long gone before the Big Event.

It's almost the same problem they (we!) have negotiating pension terms with public service unions.

And a bit like our attitude towards the environment. A species that evolved to prioritize short-term threats is pretty much doomed to ignore a long-term threat until it is too late.


Return to top