« A Natural Disaster Does Not Increased Measured GDP | Main | How did the British occupy India? »
Can a destructive storm increase measured gdp?
Say Katrina comes along and knocks down some hotels, which are then rebuilt.
We all know the "broken window fallacy" -- this sequence of events is not good for the economy. But under what conditions will it increase measured gdp?
Under one view, the money spent rebuilding the hotels would otherwise have been spent buying shoes or something else. Measured gdp should not go up. See Alex's comments below for more along these lines. (But note that Alex's fifth paragraph makes a mistake. I am not just "buying a new CD," but rather a new CD is being produced, generating income, in the analogous example he sets out, just as a new hotel is being produced to replace the old one destroyed by the storm. He doesn't come squarely to terms with how new output ever increases measured gdp. A second factual but not theoretical point is that most Katrina refugees are now earning more elsewhere.)
An alternative approach invokes the assumption of "gross substitutability," or more prosaically that new production attracts a greater expenditure than the relevant alternative. (Addendum: We also can speak of the velocity of money rising.) New production in general raises measured gdp. If a new hotel is built, why should the "gdp consequences" of that production depend on whether the lot had always been vacant, or a previous hotel on that lot was destroyed by a storm?
A further complication is that the hurricane destroys wealth. The loss of hotels induces negative income effects, which probably will lower measured gdp in other sectors of the economy. Natural disasters are not a good way to build up gdp in the longer run.
Many factors are at play. Will we consider Keynesian effects through a possible employment increase for rebuilding, or intertemporal substitution effects through a temporary boost in labor supply in repair industries? If the repairs dig into future productive capacities, short-run gdp is more likely to rise than long-run gdp.
Will natural disasters increase measured gdp in the short run, once we consider expenditure switching effects?
Your thoughts?
Posted by Tyler Cowen on October 25, 2006 at 06:00 AM in Economics | Permalink
Comments
My working assumptions is that GDP is a very gross measure, and that, yes, because it is a crude measure only of a substiture for what we are really interested in, that measured GDP will go up in the short term after natural disasters. With that said, I know very little about the actual details of the metric. Related, how do we discount things like prison construction or divorce lawyers in calculating this. I, for one, would be interested in knowing more about the areas where the measure falls apart.
Posted by: theCoach at Oct 25, 2006 8:13:06 AM
It depends on how much you like Christopher Wren.
Posted by: DK at Oct 25, 2006 8:35:28 AM
I would say there are two main effects of the storm: (1) productivity of NOLA residents drops = bad for GDP (2) increased demand for construction drives up prices in that sector (and possibly elicits extra supply) => increases measured (nominal) GDP. Net effect is probably negative.
Posted by: Thijs at Oct 25, 2006 8:40:21 AM
Don't insurance payouts play a big role? If the damaged properties are fully insured, then there is no actual loss in wealth for the owners. There is a loss for the insurers, but if their losses are not part of the local GDP calculation (if the insurance company was based in London, for example), then the disaster would trigger a massive transfer of wealth into the area.
Posted by: John at Oct 25, 2006 8:52:37 AM
What if a natural disaster destroys infrastructure in disrepair, forcing the government to rebuild it new? Because the gains of the road/port are not realized by the government (who is responsible for building it), they will underinvest until the condition of the road is REALLY bad.
Posted by: John D. at Oct 25, 2006 9:06:06 AM
Dismissing Keynesian effects and intertemporal substitution can be a big mistake. Alex's point that opportunity costs are realized in the non-disaster state of the world could be false. Consider his jazz concert example. Say the tow truck, body shop, and all other bills cost more than all the wine and cheese you were planning on buying at/after the concert. So, realistically, you may dip into your savings and spend more in the disaster period and less in future periods (intertemporal substitution).
I saw that Alex covered himself buy a saying:
"it’s conceivable that through a Keynesian effect or intertemporal substitutability of labor that GDP could rise from a natural disaster but for this to work is has to outweigh all the effects that I have listed and this is unlikely"
I think it is highly likely. If you accept that people tend to smooth income and savings, and there are imperfect credit markets, the individual(s) involved in the disaster may dip into precautionary savings during the disaster period, which precludes that from being spent in future periods. In the disaster period you will see a rise in GDP because that is the period where the precautionary savings is spent that otherwise wouldn't have been spent, because by definition it is precautionary. Alex assumes that the opportunity costs/alternatives would be realized in the same period anyway. I think this is unlikely.
I understand that the idea of precautionary savings assumes imperfect credit markets, but isn't this a fair assumption?
Posted by: Scott W at Oct 25, 2006 9:18:03 AM
1) Assets most likely to be damaged are those that have already been depreciated to some significant degree (assuming that more recent construction meets stronger building codes, and therefore were less likely to be severely damaged). Hence, these assets that are being replaced were, in some sense, less valuable. The new construction/equipment, then will appear as a boost to the GDP, if I understand the National Accounts right (and I'm not sure that I do).
2) In a like vein, consider the Y2K replacement of personal computers and major accounting programs. True, most of these systems already had fixes in place for the year becoming 2000, but the restructuring also had other significant benefits.
Posted by: Paul McMahon at Oct 25, 2006 9:24:43 AM
Broken windows (and broken hotels) increase GDP in the short term. They cause capital that was set aside for later consumption to be consumed immediately, but decrease the accumulated capital that will be used to generate future GDP.
Posted by: mobile at Oct 25, 2006 9:28:39 AM
There's also the aspect of managing risk to take into account. Could people and businesses shift their activities specifically to avoid increased risks of additional damage following a major natural disaster?
In the case of hurricanes, which have a regular, predictable season, we would see economic activity shift outside of the period of greatest risk. We would see the lowest periods of GDP coinciding with the highest disaster risk period, and the highest outside the period.
Post-disaster, you would see GDP levels oscillating much like a dampened spring - prompted by an initial dip for the disaster itself, boosted by the immediate recovery effort and insurance payouts, then with the gradually weakening oscillations as people transist from the specific risk avoidance strategy described above to more generalized risk avoidance (more "normal") strategy as time passes and the fear of a one-two punch is less a concern.
Posted by: Ironman at Oct 25, 2006 9:42:16 AM
Many people are making the mistake of thinking that if the money comes out of saving then it increases GDP. But savings are not typically held under a mattress. Higher spending coming out of savings means less investment elsewhere in the economy, i.e. less spending on capital goods elsewhere and that offset means that GDP in general doesn't rise.
Posted by: Alex Tabarrok at Oct 25, 2006 9:50:13 AM
It need not be assumed that savings are otherwise idle. The need for repair work causes the velocity of money to increase. Imagine that I woke up one morning to find that my car's gas tank was virtually drained. I would fill up with gas that morning instead of driving directly to work. Measured gdp would rise in the short run. Both my funds and my time are allocated more heavily toward the present. Of course this need not augur well for future measured gdp. But that is a simple proof of the proposition that a loss can cause measured gdp to rise in the short run.
Posted by: Tyler Cowen at Oct 25, 2006 10:00:32 AM
GDP equals hours worked times productivity.
It is a measure of production that does not
attempt to directly measure wealth or well being.
In the quarter of the storm hours worked will fall
as normal schedules are disrupted.
In the next quarter hours worked will rebound to
the previous level or trend. This will show up
in the data as a higher quarterly growth rate.
But the key question revolves around how much reconstruction
causes what would have been idle resources to become
employeed. If reconstruction draws what would have otherwise
been unemployeed resources --labor -- into productive
uses it will cause reported gdp to expand above
what would have been otherwise the case.
But this has nothing to do with the stock of wealth or
the "broken window fallacy". A storm obviously destroys
wealth and part of the increase in reported gdp is simply
a reconstruction of the destroyed wealth or the broken window fallacy.
Posted by: spencer at Oct 25, 2006 10:46:11 AM
I think Tyler's right, in principle. But the short run for one is always also the long run for another. If Tyler dips into his savings in one period to fill up his car to replace the disastrously disappeared gas, that's also a period in which somebody else tightens his belt because of savings spent in an earlier period managing the aftermath of an earlier "disaster" of a size equivalent to Tyler's. If there was a disaster of uniform size every month somewhere within the national borders, no disaster would register. I imagine this is the way it is with car crashes or house fires. Of course, that's not how it is with big natural disasters like hurricanes, so Tyler's probably still right that you could get a short-run uptick. But that doesn't effect Bastiat/Alex's general point in any interesting way.
Posted by: Will Wilkinson at Oct 25, 2006 10:52:34 AM
Yes Tyler but anyone can play the short run game. When 9/11 happened a bunch of people were instantly unemployed and GDP fell because of the wages not earned.
Posted by: Alex Tabarrok at Oct 25, 2006 11:01:24 AM
I think Tyler's gasoline example is making a common mistake
of confusing consumption with production.
GDP does not directly measure production. Rather it measures
consumption and adjust that for trade and inventories to
indirectly measure production.
The gasoline that Tyler purchases now has already been
produced and the fact that Tyler shifted it from a tank
in the ground at the gas station to a tank in his car
has no impact on gdp. He is not changing the ammount
of gasoline he is consuming. If because his gas tank was
low he had elected to walk to work -- a nonmarket event
that is not included in gdp -- his gasoline consumption
would have fallen and with a lag gasoline production and gdp would also be lower.
Posted by: spencer at Oct 25, 2006 11:17:08 AM
The broken window fallacy depends for me on the definition of wealth and the cause of the disaster.
One can see loss of wealth pretty readily with 9/11 (valuable building destroyed in act of terrorists).
When one thinks of slum clearance by hurricane, this looks more like natural destruction and a big-picture wealth-neutral thing (from a wealth - defined as healthy society - perspective).
From a micro perspective, if my 10 year old Explorer with 140,000 miles on it gets "totaled" and I get $7000 from my insurance company, why should I claim my wealth went down?
Wealth has a lot to do with perspective, so if I fallaciously believe I had no change in wealth, it could be said that my wealth did not change, yes? So it is not a fallacy that my wealth was unchanged, based on my definition of wealth?
Posted by: cfw at Oct 25, 2006 11:56:36 AM
What is "the economy" here? While I am perfectly happy using an externality adjusted measure of GDP, or something like that, I get the impression that most people are really just interested in unemployment. Couldn't we imagine a not terribly unreasonable definition of "the economy" in which "the economy" is better off but the value of the capital stock is lower?
Posted by: Scott Wood at Oct 25, 2006 12:54:29 PM
What about more qualitative effects. If economic growth is typically balanced between progress and the inertia of the status quo, what is the effect of the dislocations so many people suffer as a result of the storm? Most workers would generally be in a slightly suboptimal working position, since the costs & uncertainties of finding more optimal work are high enough to create inertia. How many workers, being forced into a disruptive state, will already be bearing those risks & costs, and use the situation, wittingly or unwittingly, to find a more optimal personal output - maybe in another city where they went to escape the storm? I wonder if there would be an immediate drop in GDP, followed by sustained gains as employment & investment were redeployed in an accelerated way.
Posted by: kebko at Oct 25, 2006 1:18:51 PM
A storm can increase GDP in the case where demolishing and rebuilding a neighborhood would be beneficial anyway. A good example is rebuilding Phoenix Park in Sacramento. Links:
http://64.233.187.104/search?q=cache:CaO1xPCbg3gJ:www.affordablehousinginstitute.org/blogs/us/2006/08/eminent_domain_13.html+fourplex+site:www.affordablehousinginstitute.org&hl=en&gl=us&ct=clnk&cd=1
http://www.shra.org/Content/CommunityDevelopment/PhoenixPark/PhoenixParkTOC.htm
Posted by: Zoran Lazarevic at Oct 25, 2006 3:13:39 PM
Hmmm I think Alex (and spencer) are right on this one though Tyler could be right in theory and wrong in practice (i.e. to get the effect you really have to stretch them assumptions).
Anyway, once upon a time I went to see Nader speak (and why not?). He was giving his usual gloom and doom picture of the US economy when somebody pointed out to him that during the Clinton years GDP per capita and even unskilled real wages rose. He answered with a big sneer and said something along the lines of "an increase in the number of car accidents increases GDP". He got a very vigorous applause from an adoring audience. Next day I used his answer as an example of bad reasoning about GDP in my Principles Macro class. We were just covering the topic.
Posted by: radek at Oct 25, 2006 3:19:57 PM
Perhaps its best to teach undergraduates that this is a good reason that glancing over at net gdp can be useful, if available. the net calculation should take into account the abnormally high level of depreciation caused by the natural disaster and give a better account of the effect of the disaster on income.
Posted by: james at Oct 25, 2006 3:37:43 PM
I may be completely off base here due to ignorance, but my intuition is that measured GDP does not really matter. The real GDP is what we will have available in the future to consume for further invest. One could argue for changes in either direction for measured GDP but I find this would be meaningless when we consider that real GDP (and presumably future living standards) may be quite different from measured GDP for a number of reasons.
Shoot this down if I've missed something key.
Posted by: Jim Outen at Oct 25, 2006 3:48:38 PM
I feel that a big storm can help the economy because it will produce jobs. There are always people that need work and when something bad happens it can pull the people together and make the economy boom. Like with the CD example, when things are destroyed then there has to be more produced with means income for the poeple in that area. So in my opinion I think that a big storm would help the GDP.
Posted by: Nick at Oct 25, 2006 4:05:37 PM
while I see many people refer to post 9/11 real gdp growth
and relate it to the broken window concept I doubt it applies.
The broken window refers to the resources used to destroyed
wealth. But the World Trade Center has not been replaced.
Yes, they repaired the Pentagon damage but that is not big
enough to impact the data.
So if you do not replace the damaged or destroyed resource
how can it be a broken window problem?
Technically, the world trade center destruction did have an
impact on gdp but it was throught the insurance part of the
equation. Real gdp calculates insurance on a net basis --
the difference between premiums and setlements. So because
9/11 lead to a large payment by insurance companies it
generated a one time upward blip in reported gdp when BEA
assumed the insurance settlement was made.
Posted by: spencer at Oct 25, 2006 4:18:04 PM
OK, time to put on my neoclassical macro hat. What we have here is a sudden decrease in the capital stock. The return to the existing capital rises, which causes an investment boom. Think of people putting in long hours rebuilding their homes and businesses and cleaning out huge amounts of muck from everywhere but the statehouse and city hall.
So we have two things at work. A decrease in productivity (Spencer's point) because of the fall in capital will cause GDP to fall. It's just harder to get stuff done when roads and buildings and communications cease to exist. But, the demand to replace these things and the resulting investment boom, and the longer hours that go with it, will cause GDP to rise. Theoretically, it's ambiguous--something of this magnitude all depends on how willing people are to put in extra hours (a "labor supply elasticity" for you econ geeks).
This is one case where GDP is a poor measure of income--this kind of touches on Jim Outen's point from a half hour ago. Literally, it is just a measure of value added in the economy due to economic activity in a given period. Net domestic product (a somewhat better measure of income) is GDP minus the fall in the capital stock from wear, tear, depreciation, and disasters. Under normal circumstances, these things track each other pretty well. But, NDP almost certainly falls as a result of Katrina, unless things are even more out of whack than they normally are in NOLA.
Posted by: Chris R at Oct 25, 2006 4:29:45 PM
