I am continuing my mini-series on modern US monetary history through the lens of my decomposition of supply inflation and demand inflation based on what I inspired by David Eagle have termed a Quasi-Real Price Index (QRPI). In this post I will have a look at the early 1980s and what have been termed the Volcker disinflation.
When Paul Volcker became Federal Reserve chairman in August 1979 US inflation was on the way to 10% and the fight against inflation had more or less been given up and there was certainly no consensus even among economists that inflation was a monetary phenomenon. Volcker set out to defeat inflation. Volcker is widely credited with achieving this goal and even though one can question US monetary policy in a number of ways in the period that Volcker was Fed chairman there is no doubt in mind my that Volcker succeed and by doing so laid the foundation for the great stability of the Great Moderation that followed from the mid-80s and lasted until 2008.
Below you see my decomposition of US inflation in the 1980s between demand inflation (which the central bank controls) and supply inflation.
As the graph shows – and as I spelled out in my earlier post on the 1970s inflationary outburst – the main cause of the rise in US inflation in 1970s was excessive loose monetary policy. This was particularly the case in late 1970s and when Volcker became Fed chairman demand inflation was well above 10%.
Volcker early on set out to reduce inflation by implementing (quasi) money supply targeting. It is obviously that the Volcker’s Fed had some operational problems with this strategy and it effectively (unfairly?) undermined the idea of a monetary policy based on Friedman style money supply targeting, but it nonetheless clearly was what brought inflation down.
The first year of Volcker’s tenure undoubtedly was extremely challenging and Volcker hardly can say to have been lucky with the timing. More or less as he became Fed chairman the second oil crisis hit and oil prices spiked dramatically in the wake of the Iranian revolution in 1979. The spike in oil prices boosted supply inflation dramatically and that pushed headline inflation well above 10% – hardly a good start point for Volcker.
Quasi-Real Price Index and the decomposition of the inflation data seem very clearly to illustrate all the key factors in the Volcker disinflation:
1) Initially Volker dictated disinflation by introducing money supply targeting. The impact on demand inflation seems to have been nearly immediate. As the graph shows demand inflation dropped sharply in1980 and the only reason headline inflation did not decrease was the sharp rise in oil prices that pushed up supply inflation.
2) The significant monetary tightening sent the US economy into recession in 1980 and this lead Volcker & Co. to abandon the policy of monetary tightening and “re-eased” monetary policy in the summer of 1980. Again the impact seems to have been immediate – demand inflation picked up sharply going into 1981.
3) Over the summer the Fed moved to hike interest rates dramatically and slow money supply growth sharply. That caused demand inflation to ease off significantly and inflation had finally been beaten.
4) The Fed allowed demand inflation to pick up once again in 1984-85, but at that time Volcker was more lucky as supply factors helped curb headline inflation.
The zigzagging in monetary policy in the early 1980s is clearly captured by my decomposition of inflation. To me shows how relatively useful these measures are and I think they could be help tools for both analysts and central bankers.
This post in no way is a full account of the Volcker disinflation. Rather it is meant as an illustration of the Quasi-Real Price Index and my suggested decomposition of inflation.
My two main sources on modern US monetary history is Robert Hetzel’s “The Monetary Policy and the Federal Reserve – A History” and Allan Meltzer’s “A History of the Federal Reserve”. However, for a critical account of the first years of the Volcker disinflation I can clearly recommend our friend David Glasner’s “Free Banking and Monetary Reform”. I am significantly less critical about money supply targeting than David, but I think his account of the Volcker disinflation clear give some insight to the problems of money supply targeting.
Martin
/ December 29, 2011Lars,
What I wondered concerning these graphs of yours, is whether, if you define demand inflation as any deviation of N from trend growth, you’re not already assuming your conclusion?
I mean when you say that ” the main cause of the rise in US inflation in 1970s was excessive loose monetary policy.”, I don’t disagree, however when you say that the graph is evidence for this view (“As the graph shows”), I have a hard time seeing how: you’ve already said the same in math when you posited the behavioral relationships between the variables.
Lars Christensen
/ December 29, 2011Martin,
It is important to not that I define demand inflation as growth in nominal GDP minus trend growth in REAL GDP. You might say that that is capturing the “business cycle” to some extent as demand inflation and the output gap by definition will be somewhat correlated. That said, I think my graphs pretty good illustrates that the measure actually quite well captures some key turning points in US monetary policy – and that is the important thing for me and that is what I try to illustrate with the comments on US monetary history.
Martin
/ December 29, 2011Lars,
When I said ‘growth’ I meant ‘real gdp’. As I said, I don’t disagree with you in defining demand inflation and supply inflation the way you do, but the graphs are not evidence for what caused what, that already follows from your definitions.
You define demand inflation, pd, and supply inflation, ps, as:
pd = n – yp, and ps = p – pd. Where yp is trend growth and is fixed.
When there is a recession therefore and y deviates from yp, the split between pd and ps is determined by p and n.
ps can be re-written as:
ps = p – n + yp or as,
ps = p – y – p + yp: ps = yp – y.
Any deviation from trend growth will show up as ‘supply inflation’. Similarly rewriting pd,
pd = p + y – yp, or pd = p – ps.
This means that whatever is not accounted for in p through deviations from yp, is called ‘demand inflation’. The sole reason therefore why you can call pd, demand inflation is because you posit that the Fed controls p through n. This is why the graphs are not evidence of a causal link.
The story you tell is very plausible, but the plausibility of your story, depends on the acceptance of the assumption by the audience.
Mike Sax
/ December 29, 2011Lars, your diagnosis-“the main cause of the rise in US inflation in 1970s was excessive loose monetary policy. This was particularly the case in late 1970s and when Volcker became Fed chairman demand inflation was well above 10%” is hardly surprising. I would have been shocked had you come to any other conclusion.
However what specifically was the policy of the Fed during the late 70s that you say was too lose? Was it what you would see as too much squeamsihness about pushing the unemploymnet rate into the double digits along with interest rates?
The “lesson” that the conventional wisdom draws from the 70s seems to be that as Friedman disproved Phillips we must prefer high unemployment to high inlfation.
If previous monetary policy was too loose it was because previous monetary policy actually took seriously the other side of the “dual mandate.”.
Niklas Blanchard
/ December 29, 2011As a note, Michael Belognia agrees with your assertion that Friedman’s k% rule for money supply growth was unfairly dismissed by a Fed that was largely hostile to Monetarism in the early 80’s.
http://www.econtalk.org/archives/2010/01/belongia_on_the.html
W. Peden
/ December 30, 2011Mike Sax,
“The “lesson” that the conventional wisdom draws from the 70s seems to be that as Friedman disproved Phillips we must prefer high unemployment to high inlfation.”
No. If Friedman was right, then we don’t have any such choice in the long-run. That is the conventional lesson drawn from the 1970s. Phillips managed to make economists waste a lot of ink on the subject of what was an optimum trade-off between inflation and unemployment, when no such trade-off existed.
Integral
/ December 31, 2011@Martin: Lars’ calculations implicitly assume a vertical LRAS and downward-sloping AD curve.
I suspect one could get qualitatively similar results by imposing an upward-sloping SRAS; the trouble then is estimating the slope of the SRAS. I did some of the math myself a week ago but haven’t finished re-deriving his indicies.