Natural disasters like floods, tornados and earthquakes destroy wealth but may spur economic production. These effects vary with the situation, especially the scale of the disaster. The destruction of wealth is a constant, but the degree to which production is spurred varies with size.
In practical terms, millions of households along the Texas Gulf coast are poorer than they were two weeks ago. In some cases, their net worth has been wiped out. But over the next months and years, a tremendous amount productive resources will go to replacing what Hurricane Harvey destroyed. These effects are indisputable.
What is in question is the degree to which such rebuilding will be a net addition to local or national output. Will we produce more goods and services overall than if Harvey had not occurred? Or will the resources expended just be shifted from another productive use, perhaps in other locations?
To think about this consider other disasters of widely varying proportions.
In 1992, an F5 tornado hit my home town of Chandler, Minn., population then about 350, as well as a swathe of farms over several miles. There were a few injuries, one of which contributed to the eventual death of an elderly woman. About a third of the houses in town were destroyed plus substantial business property and farmsteads. Damages probably totaled $10 million to $20 million.
In 1997, flooding of the Red River of the North caused extensive damage over its entire length, with the worst in Grand Forks, N.D., and surrounding area. There were no deaths, but damages, including those in Canada, were estimated at $3.5 billion.
In 2005, Hurricane Katrina damaged a very large area, but especially New Orleans. Attributed deaths vary but may have been as many as 1,836. Damages are now estimated at $108 billion. This was the most expensive natural disaster in U.S. history, though perhaps Harvey will surpass it.
In all three cases, wealth was destroyed. Useful houses, businesses and other property providing service on one day was gone or useless a day or two later. The net worth of households fell.
For society as a whole, these were losses regardless of whether owners got any payments from insurance companies or government. Such indemnities transfer money from one person or entity to another, but don’t change the total income or net worth of society overall. They do, however, play a big role in how the specific localities damaged come through in the long run.
In a famous essay, the mid-1800s French economist Frederic Bastiat noted that when a window was broken, some would observe “well, too bad for the owner, but his loss is the window repairman’s gain. There is more business.” Bastiat argued this was a fallacy. Yes, the glazier got more income. But whatever the window owner paid for repairs was money not available to buy something else. So some other shop owner or tradesman lost money. The owner had a window just as before, but did not have the new shirt or frying pan that they might have bought. Society as a whole lost a window and there was no net increase in economic activity, Bastiat argued.
He was right, at least if the assumptions of the neoclassical economics then dominant were true. But is his example of one homeowner and one tradesperson true all the way up to a massive travesty and rebuilding project that South Texas is now facing?
The town of Chandler certainly lost wealth. And there certainly was economic activity afterward. Front-loader and backhoe owners, concrete plants, block layers, carpenters, and electricians were booked solid for more than a year. Businesses in larger nearby towns sold appliances, furniture and furnishings. Farm contractors had multiple crews replacing livestock and machinery facilities. Workers had to eat and sleep somewhere, so restaurants and motels for 20 miles around had more business.
Of all common natural disasters, tornadoes have a high level of insurance coverage. So money flowed in to pay for much of the rebuilding. And there was federal money, especially for public infrastructure. These inflows of money were not enough to compensate everyone for their assets destroyed, but the amount of cash circulating was much above normal. And Bastiat was right in that the increased spending on replacing necessities did depress spending in other areas.
But on even regional or state levels, however, the impact was infinitesimal. The increased demand did not drive up the price of clothes dryers or concrete block. Businesses had good sales, but not enough to drive up their market value. Everyone understood this was temporary.
The 1997 Red River floods were bigger, but effects followed a similar pattern. There was a flurry of rebuilding. Thousands of new appliances, furnaces and electrical service panels were replaced along with acres of drywall. Dumpster services added boxes from hundreds of miles away. All local building trades were fully engaged and contractors came from extended distances in the Dakotas and Minnesota.
Here the affected area was large enough that there were increases in the prices of some high-demand items, at least temporarily. No appliance vendor put washers on sale and building materials dealers did not discount drywall. Contractors did not need to sharpen their pencils much when offering quotes. The proportion of losses covered by insurance was lower than for a tornado and the relative level of federal money higher.
Overall, the local economy was strong. Stores and other business with premises flooded out had temporary lost sales. Some went broke. But most resumed operations in remarkably short order. Here, in contrast to Chandler, while many households cut discretionary spending because of outlays on flood recovery, some others who benefitted from the boom to a point where their non-necessity purchases increased. The owner of a undamaged appliance store may have gone on a nice cruise or eaten at a steakhouse more often.
Overall, many balance sheets showed lower net worth, but current income was good. There probably was some measurable impact on North Dakota’s GDP, but nothing at the national level. Bastiat was wrong at the level of Grand Forks but still right for the nation.
Katrina, though about 50 times bigger, followed a similar pattern nevertheless. Although in this case, the scale was large enough that some observers saw national effects on some categories of goods. The pace of recovery was much slower. The scale overwhelmed FEMA, the federal government’s disaster management program. Louisiana’s state and local governments did not shine. Louisiana generally and New Orleans in particular have lower average incomes than Minnesota or North Dakota. So there was less available cash for immediate recovery spending. Much property was uninsured against flooding. And the bureaucracy took time to get money out.
Insurance claims were high enough that financial market observers noted increased bond selling by insurance companies as they converted reserves to cash to pay claims. Insurer’s profits took a hit that year. So in that sense, Bastiat again was right — the net loss of Katrina rippled beyond the increased economic activity of the recovery.
But U.S. and global financial markets are broad, deep and diversified.
Harvey may well get into the same range as Katrina. When one gets to $100 billion dollars, can’t that affect the national economy?
Here we get to John Maynard Keynes and Bastiat’s other critics. Bastiat’s generation argued increased government spending could never boost economic output because such public spending would be offset by decreased private outlays. Keynes argued that might be true in certain cases such as with full employment and no unused labor or factories.
But when there was slack in the economy with idle workers and mills, then government spending, particularly financed by central bank creation of new money, might indeed increase total output.
One could imagine a situation of a sluggish economy with laid-off carpenters and roofers, unused backhoes and piles of unsold sheet rock and air conditioners. Federal disaster spending, particularly that not offset by less spending elsewhere or by higher taxes could, in the view of Keynesians, increase total national output. Bastiat would be wrong.
Democrats historically buy this while Republicans are usually are skeptical. But now President Donald Trump and some Republicans in Congress are calling for an infrastructure spending bill at annual levels about the same as what Harvey will cost — the carrot being promised increased economic activity — putting Americans to work. So they can no longer oppose the general Keynesian principle. After the 2016 election the GOP lost concern over deficits even though this had led the Texas congressional delegation to vote against aid after Hurricane Sandy. So expect relief and recovery for Harvey to go on the national tab. Yet there is not a lot of slack in the national economy, at least by traditional measures.
How this will all play out is unclear. But we are running an interesting experiment, albeit a hard lesson to learn. Let’s hope we do not have yet another one too soon.