Thursday, January 26, 2012

The Definition of a Depression

What is a depression? What is the proper definition?

In the 19th century, people tended to use the term loosely to refer to contractions in real output often accompanied by deflation. In the Oxford English Dictionary, we get a general definition:
“5. a. A lowering in quality, vigour, or amount; the state of being lowered or reduced in force, activity, intensity, etc. In mod. use esp. of trade; spec. the Depression, the financial and industrial ‘slump’ of 1929 and subsequent years.”

(Oxford English Dictionary [2nd edn. 1989], s.v. “depression,” 5.a.).
The earliest use of the word in this sense cited in the Oxford English Dictionary is from an 1827 publication, where we read that the “commencement of the present year was marked by a continuance of that depression in manufactures and commerce, which had prevailed at the close of the preceding [year]” (The Annual Register: Or a View of the History, Politics, and Literature, of the Year 1826, 1827, p. 1).

In the 19th century, when people referred to output contractions (normally with price deflation), they spoke of a “slump in trade,” “depression of commerce” or “depression of trade and industry”, and so on. Sometimes writers spoke of a “depression” in certain particular sectors as well.

The 1870s and 1890s were widely spoken of as decades marked by depression in the 19th century, and the whole 1873–1896 period was also sometimes misleadingly referred to as a depression by contemporaries, because of the persistent price deflation in these years (even though real output growth went through several cycles).

But the sheer scale and length of the early 1930s contraction in many countries led to the expression the “Great Depression” to refer to this historically unprecedented slump.

However, today we would tend to refer to most contractions of output or downturns in the business cycle as “recessions.” A recession is often defined as two or more consecutive quarters of negative real GNP/GDP growth, accompanied by rising unemployment (Oxford English Dictionary [2nd edn. 1989], s.v. “recession,” 5.b: the earliest use in the quotations is from 1905).

The word “depression” has come mostly to refer to severe recessions. While there is no universal, formal and strictly-used definition, there is in fact a definition widely employed by economists:
“There is no formal definition of a depression, though an old joke says that a recession is when your neighbor loses his or her job, a depression is when you lose your job. An informal definition is an economic contraction in which output falls by more than 10 percent.” (Knoop 2010: 14).

“Another proposed definition of depression includes two general rules: (1) a decline in real GDP exceeding 10%, or (2) a recession lasting 2 or more years.”

“Some economists say that if gross domestic product were to decline at a 10 percent or greater annualized rate for some unspecified period of time, that would be a depression.” (Posner 2010: 218).
This definition is also used by some astute popular writers, commentators and journalists in the popular press.

I contend that the definition of a depression as an real output contraction of 10% or more is a very useful one and ought to be employed in formal economic analysis.

Recessions are those periods where output contacts by less than 10%. I think this is a useful way of categorising recessions:
(1) A mild recession would be a real GDP contraction of up to 3.33%;

(2) a moderate recession from 3.33–6.66%, and

(3) a very severe recession from 6.66–9.99%.
By these definitions, there was no “depression” in 1920–1921 by recent, revised GNP estimates: there was a moderate recession (either a 3.47% or 5.58% fall in real output).

Whatever definition of “depression” economists, bloggers or commentators on economics use, above all they ought to be consistent in their use and see where consistent use of the definition leads.

Let us take a very loose and, I charge, unsound definition of depression: simply using it to refer to the aftermath of a real output contraction where there is positive GNP/GDP growth but high unemployment.

While the “Great Depression” normally includes (1) the years after 1933 when the US economy suffered high unemployment along with (2) the actual period of contraction from 1929–1933, this is a special and often popular historical usage, and it is potential misleading: for the US had positive GNP growth after 1933 and falling unemployment until 1938 (when fiscal contraction again plunged the economy into recession).

Some Austrians claim that the US in 2010 and 2011 was, or is now (January 2012), in depression. Yet the US has positive GDP growth now and did so last year (and has had positive GDP growth since 2009). There has not been a period of actual GDP contraction since 2009. By adopting such a loose definition of depression and applying it consistently, what does this lead to? By using the same definition, I could now claim that the US was in depression for virtually the whole 1890s after 1893 because of high unemployment. I could also claim it was in depression for most of the late 1870s. Only it wasn’t really by the other important metric we have: real GNP growth. There appears to have been positive GNP growth in 1895 and 1897–1900, and after 1874 in the US. What happened was a severe shock to the economy and real GNP fell well below its potential in these years. The economy was not generating enough growth to create high employment. Thus high unemployment persisted for years. This is in fact a regular condition in capitalist economies, even though they are in periods of output expansion: full or high employment is not reached.

But this state is a different situation from an actual depression (a period of severe fall in output by 10% or more). The former is what Keynes’s (misleadingly) called an “unemployment equilibrium” (which is better called an “unemployment disequilibrium”). Keynes’s view was that real-world capitalist systems have a tendency to fluctuate around a state well below full employment:
“our actual experience … [sc. is] that we oscillate, avoiding the gravest extremes of fluctuation in employment and in prices in both directions, round an intermediate position appreciably below full employment and appreciably above the minimum employment a decline below which would endanger life.” (Keynes 2008 [1936]: 229).
But such a state (with positive GNP growth) is not a “depression”: it is better called an “unemployment disequilibrium.”

Moreover, by continuing to use the same loose definition of “depression,” I could also demonstrate that many capitalist economies outside of the 1945–1973 period were very frequently in depression, because they had high involuntary unemployment. But, by that point, I have robbed the word “depression” of useful meaning, and the same also applies to the original loose use anyway: something is wrong with this self-serving definition of “depression.” It is a rhetorical trick, unsound and ought to be discarded.

A depression can be defined with respect to severity of a real output contraction or its duration. In short, a depression is
(1) a period of actual real GNP/GDP contraction where real output falls by 10% or more, or

(2) a period of actual real GNP/GDP contraction that lasts for 2 years or more (but where real output may not fall by 10% or more).

Keynes, J. M. 2008 [1936]. The General Theory of Employment, Interest, and Money, Atlantic Publishers, New Delhi.

Knoop, Todd A. 2010. Recessions and Depressions: Understanding Business Cycles (2nd edn.), Praeger, Santa Barbara, Calif.

Posner, Richard A. The Crisis of Capitalist Democracy, Harvard University Press, Cambridge, Mass. 2010.

The Annual Register: Or a View of the History, Politics, and Literature, of the Year 1826, J. Cuthell, et al., London.


  1. Very interesting.

    A good practice might be to make this distinction as a matter of course: whenever discussing depressions, to delineate the periods of severe contraction in GNP/GDP, and the periods of high unemployment separately, since they are not coterminous.

    But your proposed definition leaves out sharp increases in unemployment, which are an important feature of the business cycle, I think. It would be nice to have as well defined a term for this as you have proposed for 'depresison'

    At any rate, it's a necessity to keep these in mind when reading any discussion of depressions, historic, or ongoing.

  2. "whenever discussing depressions, to delineate the periods of severe contraction in GNP/GDP, and the periods of high unemployment separately, since they are not coterminous."

    In case it is not clear, my definition of recession:

    "a period of actual real GNP/GDP contraction (of 2 consecutive quarters or more), generally with rising unemployment"


    "(1) a period of actual real GNP/GDP contraction where real output falls by 10% or more, generally with rising unemployment or

    (2) a period of actual real GNP/GDP contraction that lasts for 2 years or more (but where real output may not fall by 10% or more), generally with rising unemployment."

  3. I find it weird that this post isn't bombarded with comments, given that a lot of the Austrian criticism toward you deals with how 'depression' is defined

  4. In vernacular usage, an economic Depression is a period of high poverty.

    In wage-slave economies like ours, that means that a Depression is a period of high unemployment.

    I would say the 1870s and 1890s qualify, and popular usage seems to agree.