Sunday, February 2, 2014

The Recession of 1920–1921 versus the Depression of 1929–1933

The recession of 1920–1921 is alleged to have been an instance of wage and price flexibility allowing a rapid and smooth recovery from recession.

The recession lasted from January 1920 to July 1921, a period of 18 months. But an 18-month recession is relatively long by the standards of the post-1945 US business cycle, in which the average duration of US recessions fell to about 11 months.

It is also often alleged that the 1920–1921 recession shows that wage and price flexibility could have cured the depression of 1929–1933.

But one should note the following points on how the recession of 1920–1921 was quite different from the initial downturn in 1929–1930, and indeed was a highly anomalous recession in terms of its place in the economic history of the US:
(1) while the Great Depression, with its severe contraction in world trade, was virtually a worldwide phenomenon and certainly almost universal throughout the developed world, in 1920–1921 a number of nations escaped the recession: e.g., a number of European nations such as Germany, Netherlands, and Belgium and other Western offshoots such as Australia (as found in the real GDP data in Maddison 2003).

Importantly, Germany, the largest economy in Europe, was in the midst of an inflationary boom (Temin 1989: 61; Orde 2002: 146). Therefore in 1920–1921 the US was not subject to the type of shocks from collapse of world trade as in 1929–1933.

And, moreover, in 1920–1921 the US seems to have benefited from external demand from nations recovering from WWI (Gertler 2000: 242, n. 5), and in particular the boom in Germany created a strong demand for US goods in 1921 (Orde 2002: 146).

It appears, then, that the demand side of the US economy as determined by foreign demand for US exports was not badly deficient during the downturn of 1920–1921 (Kuehn 2012: 159).

(2) Although deflation in 1920 to 1921 was severe, one significant cause of the deflation was a positive supply shock in commodities due to the resumption of shipping after the war (Romer 1988: 110). After WWI, there was a recovery in agricultural production in Europe, even though American farmers had continued their production at wartime levels. When primary commodity supplies from other countries were resumed after international shipping recovered, there was a great increase in the supply of commodities and their prices plummeted. As Romer argues,
“Tiffs suggests that a flood of primary commodities may have entered the market following the war and thus driven down the price of these goods. That these supply shocks may have been important in stimulating the economy can be seen in the fact that the response of the manufacturing sector to the decline in aggregate demand in 1921 was very uneven …. The industries that were most devastated by the downturn were those in heavy manufacturing …. On the other hand, nearly all industries… that used agricultural goods or imports as raw materials experienced little or no decline in labour input in 1921 .... That industries related to agricultural goods and imports flourished during 1921 suggests that beneficial supply shocks did stimulate production in a substantial sector of the economy” (Romer 1988: 111).
Vernon (1991) comes to the same conclusion as Romer: the deflation in 1920-1921 was caused not just by a decline in aggregate demand but also by a positive aggregate supply shock.

(3) in 1920–1921 US consumption behaved very differently to what happened in 1929–1930.

As we can see below in the graph index of the consumption component of US real GDP, real consumption spending fell sharply in 1929–1930, but actually rose from 1920–1921 (Gertler 2000: 242; data from Cole and Ohanian 2000: 185, Table 1).

Gertler (2000: 242) argues that the end of WWI released pent-up demand for consumption goods that continued in 1920–1921.

So the forces of debt deflation and high real interest rates were offset by the postwar consumption demand (Gertler 2000: 242).

Even real private investment did not slump as much in 1920–1921 as it did from 1929–1930, as we can see in the graph below.

Since we have already seen that neither (1) export demand nor (2) US consumer demand appears to be a major cause of the recession of 1920 to 1921, it follows that a slump in US domestic investment was the primary problem.

But what exactly did the fall in private investment represent in 1920–1921 and what caused it?

To that, let us turn to (4).

(4) Temin’s analysis of the causes of the recession of 1920–1921 is very interesting:
“The decline that started in 1929 was due to a failure of aggregate demand … . The depression of 1920–21, by contrast, was due largely to a shift of demand. The war had ended, and demobilization resulted in a massive transfer of demand among industries and firms. In the United States, government expenditures contracted sharply in 1920, reducing demand and releasing workers. But the war had suppressed private demand and increased private wealth. The United States had borrowed from its citizens and loaned to allied governments, making a rapid transition from international borrower to international lender. As a result private demand rose to take the place of war expenditures. This wealth effect made for a short depression after the war. The comparable effect was strong enough to obviate any recession after the Second World War.” (Temin 1989: 60).
According to Temin, then, this was essentially a post-war reconstruction recession: this type of recession is qualitatively different from those characterised by severe failures of aggregate demand and shocks to business expectations.

Although I have not yet looked in greater detail at Temin’s explanation and I would not strongly commit to that view without further research, it at least deserves consideration.

And I would also note that the slump in investment demand was presumably also partly induced by the contractionary monetary policy implemented by the Federal Reserve before 1920. I also note in passing that Paul A. Samuelson had some interesting analysis of the recession of 1920–1921 (Samuelson 1943: 47–50) and attributed it to the collapse of a boom in 1919 that was itself fuelled by continued (but falling) large government spending, business inventory accumulation and price inflation.

(5) the levels of US private debt were lower in 1921 than the very high levels in 1929.

Given the lower levels of US private debt in 1920–1921 and the fact that demand and business confidence revived in 1921, the US escaped a cumulative process of factors like that which caused the severity of the depression from 1929–1933.

While there was a debt deflationary effect in 1920–1921, one must remember that the deflation of 1920–1921 came after the high inflation of the First World War and 1919 boom which must have inflated away the real value of the private debts of many people contracted in the years before 1920 (an important point made by Kuehn 2012: 159).

So the deflation of 1920–1921 was not nearly so severe when one considers the preceding inflation (Kuehn 2012: 159), and the likelihood that people expected a post-war deflationary episode (Bordo, Erceg, and Evans 2000: 235–236).

In contrast, the massive rise in US private debt in the 1920s occurred with a relatively stable price level and low inflation (as noted by Kuehn 2012: 159), and people were not expecting a severe protracted deflation in 1929.

And, as we have seen, in 1929 the level of private nominal debt was much higher than in 1920 (Dimand 1997: 140).

One must also look at the composition as well as higher level of private debt in 1929, as Irving Fisher noted:
“From 1921–29, as the boom developed, the new corporate issues took more and more the form of stocks instead of bonds. This policy of reducing the proportion of bonds had one good effect: It left the corporations less encumbered with debt so that, despite the depression, many corporations kept in a strong position throughout the whole of the depression. This advantage, however, was more than offset by shifting the debt burden from the corporations to the stockholders. That is, in order to buy the stock, many persons borrowed, so that, instead of being indebted collectively in the form of a corporation, they became indebted individually. Moreover, their borrowing was of the most dangerous type: largely margin accounts with brokers, whose loans were call loans. Thus, upon the corporate equities represented by common stocks was superimposed a structure of equities represented largely by margin accounts and brokers’ loans.” (Fisher 1933: 72).
So, although there was debt deflation in 1920–1921, the lower levels of private debt and previous inflation meant that its effects were not as bad as in 1929–1933, and fundamentally with the revival in investment and business confidence in 1921, the cumulative effect of debt deflation (along with a host of other factors) was stopped in its tracks as the recovery proceeded in 1922 and later years.

(6) The flexibility of US real and nominal wages in the period from 1914 to 1921 – and especially the downwards flexibility in 1920 to 1921 – stands out as an historically unprecedented event in US economic history (Sundstrom 1992: 433): it is a deviation from a general trend of relative wage rigidity. Even the late 19th century showed an increasing and significant relative wage stickiness, so that, at least as far back as the 1880s/1890s, there was no golden age of wage flexibility before 1929 that was somehow destroyed by government intervention.

The fall in nominal and real wages in 1920–1921 was mostly the consequence of the fact that they had been driven to high levels in the First World War. That this wage adjustment helped the US economy in this atypical instance, especially in the export sectors, is no doubt true, but it does not prove that wage flexibility is an appropriate or reliable solution to involuntary employment in other normal circumstances when recessions occur.

(7) the 1920 to 1921 recession saw no disastrous financial crisis. Although some bank failures occurred, there were no mass bank runs and collapses as in 1929–1933 (Brunner 1981: 44).
All in all, the recession of 1920–1921 was obviously a highly anomalous downturn, and merely because there was a recovery in 1921, this does not prove that the recovery was caused by wage and price flexibility, given
(1) that the recession was not fundamentally caused by deficient aggregate demand in the GDP components of (a) export demand and (b) US consumer demand;

(2) that positive supply-side shocks were important;

(3) that there was a rise in demand for American exports in 1921, and

(4) the possibility that the downturn was a reconstruction recession.
All of these suggest that the recovery in 1921 can be understood as a demand-side phenomenon.

Further Reading
“The US Recession of 1920–1921: Some Austrian Myths,” October 23, 2010.

“There was no US Recovery in 1921 under Austrian Trade Cycle Theory!,” June 25, 2011.

“The Depression of 1920–1921: An Austrian Myth,” December 9, 2011.

“A Video on the US Recession of 1920-1921: Debunking the Libertarian Narrative,” February 5, 2012.

“The Recovery from the US Recession of 1920–1921 and Open Market Operations,” October 4, 2012.

“Rothbard on the Recession of 1920–1921,” October 6, 2012.

External Links
David Glasner, “Daniel Kuehn Explains the Dearly Beloved Depression of 1920–21,” Uneasy Money, February 1, 2012

Murphy, Robert P. “Krugman and Kuehn take me to the Woodshed on the 1920–1921 Depression,” Free Advice, 24 January 2012

Daniel Kuehn, “Glasner on 1920–21,” Facts and Other Stubborn Things, February 2, 2012

Daniel Kuehn, “Krugman on 1920–21,” Facts and Other Stubborn Things, January 23, 2012

Daniel Kuehn, “My CJE article has been published,” Facts and Other Stubborn Things, January 17, 2012

Daniel Kuehn, “Casey Mulligan would have loved the 1920–1921 Depression,” Facts and Other Stubborn Things, August 20, 2011

Daniel Kuehn, “Ryan Murphy on my 1920–21 Paper,” Facts and Other Stubborn Things, April 5, 2011

Daniel Kuehn, “Krugman on the 1921 Depression,” Facts and Other Stubborn Things, April 1, 2011

Bordo, Michael, Erceg, Christopher and Charles Evans. 2000. “Comment” (on “Re-Examining the Contributions of Money and Banking Shocks to the U.S. Great Depression”), NBER Macroeconomics Annual15: 227–237.

Brunner, K. 1981. The Great Depression Revisited. Nijhoff, Boston and London.

Cole, Harold L. and Lee E. Ohanian. 2000. “Re-Examining the Contributions of Money and Banking Shocks to the U.S. Great Depression,” NBER Macroeconomics Annual 15: 183-227

Dimand, Robert W. 1997. “Debt-Deflation Theory,” in D. Glasner and T. F. Cooley (eds.), Business Cycles and Depressions: An Encyclopedia, Garland Pub., New York. 140–141.

Fisher, Irving. 1933. Booms and Depressions: Some First Principles. George Allen and Unwin, London.

Gertler, Mark. 2000. “Comment,” NBER Macroeconomics Annual 15: 237–258.

Kuehn, Daniel. 2012. “A Note on America’s 1920–21 Depression as an Argument for Austerity,” Cambridge Journal of Economics 36.1: 155–160.

Maddison, Angus. 2003. The World Economy: Historical Statistics. OECD Publishing, Paris.

Orde, Anne. 2002. British Policy and European Reconstruction after the First World War. Cambridge University Press, Cambridge.

Romer, C. 1988. “World War I and the Postwar Depression: A Reinterpretation based on alternative estimates of GNP,” Journal of Monetary Economics 22.1: 91–115.

Samuelson, Paul A. 1943. “Full Employment after the War,” in Seymour E. Harris (ed.), Postwar Economic Problems, McGraw-Hill, New York and London. 27–53.

Sundstrom, William A. 1992. “Rigid Wages or Small Equilibrium Adjustments? Evidence from the Contraction of 1893,” Explorations in Economic History 29.4: 430–455.

Temin, P. 1989. Lessons from the Great Depression. MIT Press, Cambridge, MA.

Vernon, J. R. 1991. “The 1920–21 Deflation: The Role of Aggregate Supply,” Economic Inquiry 29: 572–580.


  1. something interesting from Krugman

  2. Yeah, that's interesting. Argentina purportedly has a flexible exchange rate, but their CB intervenes quite a lot. They still have around 6% unemployment though. Does that comport with a supply side constraint?