VoxEU Column Economic history Global crisis Monetary Policy

Lessons from money growth and inflation in the US since 2008 and the post-WWI German hyperinflation

The quantity theory of money implies that sustained inflation requires a sustained increase in the money supply. It does not, however, imply that the reverse is also true. This column explores and illustrates this issue by comparing inflation in the US following the collapse of Lehman Brothers with Germany’s hyperinflation experience after WWI. A key factor explaining the vastly different inflation experiences is how the monetary expansion translated into demand. The Fed’s base expansion did not translate into demand for goods and services, whereas the German monetary expansion was motivated by the government’s hunger for seigniorage revenues.

The famous 1963 dictum “Inflation is always and everywhere a monetary phenomenon”, by Milton Friedman succinctly summarises a basic implication of the quantity theory of money for the relation between money and prices. It has been an empirical beacon for generations of students of inflation as well as for central bankers. Translated into more precise terms, it implies that a necessary condition for sustained inflation is a sustained increase in the quantity of money. But it does not imply that all persistent increases in the quantity of money are necessarily inflationary. In particular, when increases in money supply are matched by increases in money demand even the simple quantity theory implies that the price level should not change.  More generally whether persistent monetary expansion induces persistent inflation depends on a number of additional economic and institutional factors that transcend the run of the mill quantity theory of money.1  

This column illustrates and discusses those issues, first by documenting the dramatic difference between US inflation since Lehman’s collapse and inflation during the first half of the post-WWI German hyperinflation for identical rates of expansion of high-powered money, and second, by analysing the reasons for this difference.2

Figure 1 displays the evolution of high-powered money and inflation in the US starting in September 2008 till September 2014 and in Germany starting from December 1920.3 The values of the monetary bases and of the price levels in both the US and Germany are normalised to 100 at the beginning of each of those two periods in order to provide a common comparative scale for the two episodes.4 For the same reason, the initial periods of the two episodes are located at the same extreme left-hand sides of the horizontal axis, where the chronological dates for the US are displayed on the lower horizontal axis and those for Germany on the upper horizontal axis.

Between September 2008 and September 2014 base money in the US increased by a factor of 4.35 (435%). In order to compare the inflationary consequences of the same base money expansion in today’s US with those of the German hyperinflation 90 years ago, the German data is truncated when the cumulative rate of base money expansion equals that of the US between September 2008 and September 2014. This occurs in September 1922, which is about 15 months prior to the end of the hyperinflation. The figure essentially replaces chronological time with time units anchored on identical rates of base money expansion. The blue and red lines in Figure 1 refer to the US and Germany respectively. The solid lines stand for the evolutions of the base money stocks and the dashed lines for the evolution of the price levels, all in comparison to their respective base periods. Consequently a point on any of the curves shows by how much high powered money or the price level have increased in comparison to their common base period.

Figure 1 

The behavior of the monetary base and the price level in the US since Lehman's collapse and during the German hyperinflation: A comparison

Notes: The values of the monetary bases and of the price levels in the US and Germany are all are normalized to 100 at the beginning of each of the two periods (Sept 2008 for the US and December 1920 for Germany).
Sources: (i) H and CPI for USA: Federal Reserve Bank of St Louis Data Base; (ii) H and CPI for Germany: Calculated from data in Table 1 of Cukierman (1988).

For Germany the figures show that, following a period of about seven months during which the price level increased less than high-powered money, there was a persistent acceleration of inflation much beyond the rate of base money expansion. As a consequence, the German price level in September 1922 was 24 times higher than in December 1920. During the same period, base money increased only by a factor of 4.35. By contrast, in the US the cumulative rate of increase in the price level is consistently much lower than the cumulative rate of base expansion. The cumulative CPI increase between Lehman’s collapse and September 2014 is 12.4%. This is obviously miniscule in comparison to the 435% increase in the monetary base.

What are the reasons for this dramatic difference in inflation outcomes? The most important economic reason is that, in post Lehman’s US, expansion of the base was hardly translated into higher demands for goods and services, while in Germany during the twenties practically all the expansion in high-powered money was used from the start by Government to finance the state budget.5 In the US since September 2008 about three quarters of the huge monetary base expansion took the form of an increase in bank reserves at the Fed without any appreciable impact on credit growth. As a consequence, higher order monetary stocks in the public’s portfolio and (relatedly) the transmission to the demand for goods and services was much weaker than suggested at first blush by the figures on base expansion.6 The upshot is that the Fed’s base expansion did not translate into demand for goods and services whereas the German monetary expansion was motivated from the start by a strong hunger on the part of government for seigniorage revenues.

Lurking behind this are important differences in monetary institutions. The Fed is largely independent from political authorities and committed to a low rate of inflation. By contrast, the Reichsbank (the German central bank during the hyperinflation) was totally under the control of German political authorities. For political reasons – related to the structure of war reparations imposed on Germany, in conjunction with a post-war damaged tax collection apparatus – German political authorities had a major incentive to heavily rely on the printing press.7 This difference is critical for the anchoring of inflationary expectations. As highlighted by the New-Keynesian literature, the behaviour of those expectations has a first order effect on price adjustments in the economy, and therefore on the rate of inflation.8

In addition, when inflationary expectations go up, the speed of price adjustment by firms in the economy goes up as well after a while. This mechanism further reinforces the acceleration in the rate of inflation.9 This process reached its full impact on inflation in Germany during the second half of the hyperinflation. During the German hyperinflation central bank actions reinforced a trend of increase in the velocity of circulation of money (Cagan 1956). By contrast, in the US since Lehman’s Collapse, the low interest policy of the Fed reduced the velocity of circulation. As previously explained those differences are traceable to differences in the origins of the original shocks along with different institutional setups.

During the German inflation there was no anchor for expectations. As a consequence, as inflation picked up, those expectations ultimately adjusted upward which raised inflation further. By contrast, in today’s US, expectations are tightly anchored by the following two institutional devices:

  1. Only the Fed decides on monetary policy and the Fed is committed to an explicit inflation targeting regime; and,
  2. Relatedly, the US Government is prohibited from relying on seigniorage to finance deficits.

Admittedly the Fed has to turn the profits that accrue to it as a result of its independent monetary operations to government every week. But government cannot influence the size of those profits in order to tailor them to its fiscal needs. Those two factors contribute a lot to the current credibility of monetary policy in the US and through it to the anchoring of inflationary expectations.

Last but not least, following WWI, Germany had little or no access to international capital markets. As a consequence, the main (if not only) way to finance deficits was via monetary expansion. By contrast the US enjoys unparalleled access to both home and international capital markets as well as the privilege to borrow in its own currency. Thus, US fiscal authorities have no reason to rely on seigniorage revenues even in the face of substantial deficits. Consequently, the credibility of low US inflation is backed not only by the law that prohibits government from directly borrowing from the Fed but, more fundamentally, by the US Government’s easy access to financial markets.

The inflationary experience of the post-WWI German hyperinflation contributed a lot to monetary policymakers belief that excess liquidity creation is likely to be inflationary. As a matter of fact, during the first several years following the outbreak of the Global Crisis, some practically oriented central bankers worried that the large liquidity levels created by the Fed will ignite the fires of inflation. An important lesson from the comparison between German inflation some 93 years ago with recent inflation in the US is that, due to profound differences in economic circumstances and institutions, such fears are grossly exaggerated.  

References

Cagan, P (1956) “The monetary dynamics of Hyperinflation”, in M Friedman (ed), Studies in the Quantity Theory of Money, University of Chicago Press, Chicago, Il.

Cukierman, A (1988) “Rapid inflation – Deliberate policy or miscalculation?”, Carnegie Rochester Conference Series on Public Policy, 29.

Cukierman, A (2015) “US banks’ behavior since Lehman’s collapse, bailout uncertainty and the choice of exit strategies”, CEPR Discussion paper, DP10349.

Friedman, M (1963) Inflation causes and consequences, Asian Publishing House,

Gali, J (2008) Monetary policy, inflation, and the business cycle, Princeton University Press, Princeton, NJ.

Woodford, M (2003) Interest and prices, Princeton University Press, Princeton, NJ

Endnotes

[1] More sophisticated versions of the quantity theory accommodate some but not all of those factors. An early example is Cagan (1956) who explicitly recognises the effects of changing expectations on velocity and the dynamics of inflation.

[2] This column draws on sections 5 and 6 of Cukierman (2015).

[3] Cagan (1956) locates the beginning of the German hyperinflation in December 1920. 

[4] Hence, by construction all four graphs start from a common base of 100.

[5] In Cukierman (1988: 47), I calculate that during 1921, 1922, and 1923, seigniorage financed 56%, 64%, and 89% of the German Government budget, respectively.

[6] Details appear in Figures 1 and 6 of Cukierman (2015).

[7] A detailed discussion appears in section 7 of Cukierman (1988).

[8] Standard references on New-Keynesian models are Woodford (2003) and Gali (2008).

[9] In terms of the New-Keynesian model this means that the Calvo coefficient changes with the customary level of inflationary expectations.   

6,509 Reads