Aug 1 2009, 4:57PM
When Does a Depression or a Recession End?
Economists' joke: A recession is when your neighbor loses his job; a depression is when you lose your job.
The point of the joke is that neither "recession" nor "depression" is well defined and so the line between them is indistinct to the point of vanishing. (Economists like to joke about depressions and recessions because they are untouched by them, especially if they have tenure.) Until sometime after World War II, all substantial economic downturns, of which there had been many, were referred to as "depressions." The 1930s depression was the gravest, at least in modern times, and so came to be called the Great Depression. Oddly, when that happened, lesser depressions, past and future, became labeled (or relabeled, in the case of the earlier depressions) "recessions." What we now call "World War I" was called before the second world war "the Great War." Yet, having decided that the 1914-1918 war was the greatest war in history in cost, including number of deaths and gravity of political consequences, the word "war" was not retired as the name of earlier conflicts. People didn't start saying "the American Civil Fight" or the "Napoleonic Fights." Yet nowadays anything that doesn't measure up to the Great Depression in terms of GDP decline or unemployment decline or some other measure of economic loss is called a "recession," including the current economic downturn--which, because of its severity (unequaled since the Great Depression) some have started calling "the Great Recession."
All this shows is that Americans' use of language is debased; and we knew that, and that I am a language fusspot. What is more interesting and important is how the media and the economics profession define a recession. (There is no accepted definition of a depression any more, except "comparable to the Great Depression.") The media define it as two consecutive quarters in which Gross Domestic Product (the market value of all goods and services sold in the economy) falls, which is crude but serviceable, except that it doesn't enable the beginning of the recession to be pinpointed to a month. The National Bureau of Economic Research uses a similar measure but looks at other economic indicators besides GDP, such as unemployment, but cannot "call" a recession when it starts because protraction is one of the criteria. It took it about a year to decide that the current "recession" (I call it a "depression," because of its fiscal and political consequences, which are looming as enormous) began in December 2007.
But the really important question is when a recession ends. The media regard it as ending when GDP stops falling; the economists when it starts rising. Both definitions are misleading, as the statistics of the current situation show.
For simplicity, assume that GDP in 2007 was 100. In 2008 it was less than four-tenths of one percent greater: hence 100.4. In the first quarter of 2009, it fell at an annual rate of 6.4 percent: that is, it declined by 1.6 percent. In the second quarter, just ended, it declined at an annual rate of 1 percent, which means that it fell .25 percent that quarter. Hence, by the end of July, GDP was 98.55, compared to 100 in 2007. Suppose it is flat in the third quarter of this year and rises at an annual rate of 1 percent in the fourth quarter (i.e., it rises by .25 percent--approximately; I am doing some minor rounding). (I am not forecasting; these are hypothetical numbers.) Then GDP for 2009 as a whole would be 98.8. That looks like a small decrease since 2007. But this ignores the GDP trend line. GDP grows at an inflation-adjusted rate (all my numbers are inflation-adjusted) of about 3 percent a year on average. Hence GDP in 2008 "should" have been 103 and 106 in 2009. At 98.8, therefore, it would be 7.2 percent below trend. Nonetheless, most journalists, economists, and government officials would say, given my numbers, that the recession had "ended" in either the third or fourth quarter of 2009.
By the same token, the Great Depression ended in 1933, when GDP began to rise, though when it began GDP was a third below its 1929 level and unemployment was at 25 percent.
It seems to me that a better definition--one that would give a more realistic picture of the business cycle--would be that a recession (or depression) ends when GDP returns to (or near) its trend line. Until that happens, the economy is in trouble and measures to speed recovery should continue to be considered. Otherwise, when GDP begins to grow, however modestly--or even when it just stops falling--people will say: the recession is over, so let's forget about the economy for a while, even if unemployment is still growing, foreclosures are increasing, defaults and bankruptcies are increasing, and, in short, the economy is performing in a completely unsatisfactory manner. Maybe nothing can be done at that stage but let economic "nature" take its course; but complacency and false optimism should be avoided.
(Photo credit: http://commons.wikimedia.org/wiki/File:Actual_potential_GDP_output_gap_CBO_Jan_09_outlook.png#file)
Comments (12)
Comments on this entry have been closed.
Richard A. Posner
To the average unemployed worker, it is irrelevant.
To Wall Street it is equally irrelevant.
I'm less concerned with the "end" of the recession in terms of the bottom real-GDP figure than I am with the lack of an engine for future growth. Right now even the mediocre improvement in production we're witnessing relies heavily on government deficit spending. The point of such spending is to restore confidence ("animal spirits") and to curtail job losses.
But the risk of such spending is that if the pre-crash psychology isn't restorable and shifts in consumer behavior turn out to be of long duration, in which case "stimulus" spending only pulls forward future demand. The magic day when we hit the magic number of "0" isn't meaningful if, instead of reflecting real, fundamental improvement in the economy, it was only accomplished by a fiscal distortion of the data which in turn merely temporarily masks the underlying ill health of the private economy.
As an example - the "cash for clunkers" program. The government has essentially offered large grants to the owners of older, fuel-inefficient, low value (less than $4,500 trade-in) but nevertheless fully functional vehicles, if they will buy a new, fuel-efficient vehicle, and destroy the engine and transmission of their old one. The preferred method is to pour sodium-silicate in the oil tank and run the engine at high-RPM until it is damaged beyond further use.
Now, these are old engines, probably almost all over a decade old. My 12-year old Chrysler sedan, for example, was "too new", given 1998's improved CAFE standards, and thus did not qualify for this program. The average remaining life of these clunker engines was probably between 1 and 3 years anyway.
My point is: At some point soon these cars would have broken down and would have had to have been replaced anyway, and their owners would probably have done so even without free money from the government. Replacing old, worn out cars is the major driver of demand for new car production.
The "clunkers" program, therefore, may just serve to accelerate that future demand - which is just a way of robbing Peter to pay Paul. What are we going to do in a year or two when car sales are "mysteriously" a few hundred-thousand below expectations? Stimulate again so GDP doesn't look like it's contracting? The economy would start to become utterly dependent on "stimulus" just to tread water - in the same way a long-addicted smoker needs constant drags on a cigarette, not to feel better, but just to maintain normalcy.
The same goes for most of the deficit spending that is going on right now. If the fiscal stimulus spends huge amount of money to prop up GDP (all the way to 0!), but that spending does not accomplish the goal of changing investors' moods and consumers' behaviors, then it merely brought forward government demand by means of debt which has to be paid for eventually out of future consumption. Again, robbing Peter to pay Paul.
Without changing behaviors, without being able to get there in the absence of huge and extraordinary government expenditures, and without any obvious future "growth engine" in the economy to remedy the huge amount of unemployment, getting to 0 doesn't seem to be such a cheery occasion.
We seem to have forgotten the psychology behind a working economy. People must believe that everything is fine to behave as such. I'm not saying that the "normal" behavior up to this event should come back (credit means get it was and always will be a terrible mantra), but people are hoarding because the flickering box in the living room tells them that their money isn't safe yet. Vehicle sales weren't driven by the breakdown of machinery and it's replacement, they were driven by fashion. Cars were replaced every three years or so because of vanity and easy loan approval, not because people put forty thousand or so city mile per year on a vehicle.
Well, cars are traded in tiers from those that get new cars every other year to those who buy cars only a year away from the scrap yard. The frequency at which new-car buyers are able to replace (and therefore, the number of new-car sales) is affected by higher or lower resale values for their used cars, which is related to the depreciation and scrappage rates. So, perhaps new car sales are not "caused" by the need to replace worn out used cars - but they are strongly affected by that need indirectly through the market chain.
A good measure of the magnitude of this relationship would be to compare scrappage rates with new car sales - or equivalently - the net annual addition to registered automobiles as compared to new car sales. From year to year, it is fairly chaotic, but on average what you would see for recent years are scrappage-to-sales ratios of about 80-90%, and recently, of over 100% - which means we are currently demolishing more worn-out old cars than we are building and selling.
From these data, I conclude that the need to replace is a primary driver of new car sales, and will become even more significant a factor in the next few years if the economy experiences a substantial recovery.
its very nice to read about the economy by well paid, employed, smart economists, but what about the people who are unemployed and under-employed?! it should be noted that a real recovery of the 30's depression occured when the threat of war was recognized...and not much before that time. bottom line...recovery means lots of jobs and this recovery will be less jobs! the present wars in Iraq and Afghanistan will be over one day soon and its positive effects on the economy will end. some thing to factor into the recovery equation. instead of observations and predictions, a study should be undertaken on how to restructure economies in the times to come that will deal with high energy costs, reduced employment and state sponsored threats. end of speech!
"[T]he very influences that have created ... flexibility and precision have also allowed -- indeed encouraged -- many writers to produce prose that is quite bad."
- Joseph M. Williams
Since you are a "language fussbudget," I'm sure you will appreciate my pointing out that it is not Americans' use of language that is "debased," but economists who debase the language surrounding the recovery.
Come down from your ivory tower and you'll realize there isn't a single person on "Main Street" who defines "recovery" as jobless, or ties it to GDP. Most people outside of Washington and Wall Street define "recovery" as becoming gainfully employed in a job that will show real increases in income over time, not in a job that provides stagnant wages over decades.
GDP trend lines are irrelevant if millions of Americans continue to remain un- and under-employed.
Isn't that the point that Posner was trying to make? That the ivory tower economists aren't using "recession" and "depression" in ways that really reflect the impact the economy is having on ordinary people's lives?
Got hung up on this statement in Posner's story: "It seems to me that a better definition--one that would give a more realistic picture of the business cycle--would be that a recession (or depression) ends when GDP returns to (or near) its trend line."
That's why it seemed Posner's up there with the rest of them in the ivory tower... To me, the economic catastrophe ends when the millions of Americans dumped from their jobs or are now underemployed get to earn a real living wage again.... If the GDP floats up toward its trend line once again and millions of people remain underemployed, we're not recovered as a nation.
Very interesting take on the question whether or not the recession is over. Unfortunately, the recession has affected most of us and it's too soon for the recession to be over. Here's more discussion about this.
Is there any correlation between the return of economic growth to previously-expected levels and job growth? The media always claims that job losses tend to continue after a recession is technically over, but I wonder if it's because they are using an inaccurate assessment of "over."
There are various measures of the polemics about "recession's". The easiest and most common is 2 consecutive quarters of negative GDP growth. The problem with that is that GDP growth (or loss) is measured 2 quarters (six months) after the event.
By the same token the end of a recession is not known until numbers for growth have been established. So the reporting is accurate, no one knows the end of the recession, or the beginning, and thus job losses continue. Once the contraction abates, economic growth as measured by GDP will begin to be reported.