Money and Inflation

In light of comments on my post about inflation, e.g. that deficits and supply shocks cannot produce a long-term inflation, etc., let me try to convince you that at least one member of the Fed thinks about inflation exactly as I wrote. This is from chapter 24 of Federal Reserve Governor Frederic S. Mishkin's textbook The Economics of Money, Banking, and Financial Markets (8th ed.), graphs and all. The chapter is called "Money and Inflation":

Meaning of Inflation

You may have noticed that all the empirical evidence on the relationship of money growth and inflation discussed so far looks only at cases in which the price level is continually rising at a rapid rate. It is this definition of inflation that Friedman and other economists use when they make statements such as "Inflation is always and everywhere a monetary phenomenon."

This is not what your friendly newscaster means when reporting the monthly inflation rate on the nightly news. The newscaster is only telling you how much, in percentage terms, the price level has changed from the previous month. For example, when you hear that the monthly inflation rate is 1% (12% annual rate), this indicates only that the price level has risen by 1% in that month. This could be a one-shot change, in which the high inflation rate is merely a temporary, not sustained. Only if the inflation rate remains high for a substantial period of time (greater than 1% per month for several years) will economists say that inflation has been high.

Accordingly, Milton Friedman's proposition actually says that upward movements in the price level are a monetary phenomena only if this is a sustained process. When inflation is define as a continuing and rapid rise in the price level, almost all economists agree with Friedman's proposition that money alone is to blame.

Views of Inflation

Now that we understand what Friedman's proposition means, we can use the aggregate supply and demand analysis learned in Chapter 22 to show that large and persistent upward movements in the price level (high inflation) can occur only if there is a continually growing money supply.

How Money Growth Produces Inflation

First, let's look at the outcome of a continually growing money supply (see Figure 2). Initially the economy is at point 1, with output at the natural rate level and the price level at P₁ (the intersection of the aggregate demand curve AD₁, and the short-run aggregate supply curve AS₁). If the money supply increases steadily over the course of the year, the aggregate demand curve shifts rightward to AD₂. At first, for a very brief time, the economy may move to point 1' and output may increase above the natural rate level to Y', but the resulting decline in unemployment below the natural rate level will cause wages to rise, and the short-run aggregate supply curve will quickly begin to shift leftward. It will stop shifting only when it reaches AS₂, at which time the economy has returned to the natural rate level of output on the long-run aggregate supply curve.¹ At the new equilibrium, point 2, the price level has increased from  P₁ to P₂.

click on figure to enlarge

If the money supply increases the next year, the aggregate demand curve will shift to the right again to AD₃, and the short-run aggregate supply curve will shift from  AS₂ to AS₃; the economy will move to point 2' and then to point 3, where the price level has risen to P₃. If the money supply continues to grow in subsequent years, the economy will continue to move to higher and higher price levels. As long as the money supply grows, this process will continue, and inflation will occur. High money growth produces high inflation.

Can Other Factors Besides Money Growth Produce a Sustained Inflation?

In the aggregate demand and supply analysis in Chapter 22, you learned that other factors besides changes in the money Supply (such as fiscal policy and supply shocks) can affect the aggregate demand and supply curves. Doesn't this suggest that these other factors can generate high inflation? The answer, surprisingly, is no. To see why high inflation is always a monetary phenomenon, let's dig a little deeper into aggregate demand and supply analysis to see whether other factors can generate high inflation in the absence of a high rate of money growth.

Can Fiscal Policy by Itself Produce Inflation? To examine this question, let's look at Figure 3, which demonstrates the effect of a one-shot permanent increase in government expenditure (say, from $500 billion to $600 billion) on aggregate output and the price level. Initially, we are at point 1, where output is at the natural rate level and the price level is P₁. The increase in government expenditure shifts the aggregate demand curve to AD₂, and we move to point 1', where output is above the natural rate level at Y_1'. Because of this, the short-run aggregate supply curve will begin to shift leftward, eventually reaching AS₂, where it intersects the aggregate demand curve AD₂, at point 2, at which output is again at the natural rate level and the price level has risen to P₂.

click on figure to enlarge

The net result of a one-shot permanent increase in government expenditure is a one-shot permanent increase in the price level. What happens to the inflation rate? When we move from point 1 to 1' to 2, the price level rises, and we have a positive inflation rate. But when we finally get to point 2, the inflation rate returns to zero. We see that the one-shot increase in government expenditure leads to only a temporary increase in the inflation rate, not to an inflation in which the price level is continually rising.

If government spending increases continually, however, we could get a continuing rise in the price level. It appears, then, that aggregate demand and supply analysis could reject Friedman's proposition that inflation is always the result of money growth. The problem with this argument is that a continually increasing level of government expen­diture is not a feasible policy. There is a limit on the total amount of possible govern­ment expenditure; the government cannot spend more than 100%, of GDP. In fact, well before this limit is reached, the political process would stop the increases in government spending. As revealed in the continual debates in Congress over balanced budgets and government spending, both the public and politicians have a particular target level of government spending they deem appropriate; although small deviations from this level might be tolerated, large deviations would not. Indeed, public and political perceptions impose tight limits on the degree to which government expenditures can increase.

What about the other side of fiscal policy-taxes? Could continual tax cuts generate an inflation? Again the answer is no. The analysis in Figure 3 also describes the price and output response to a one-shot decrease in taxes. There will be a one-shot increase in the price level, but the increase in the inflation rate will be only temporary We can increase the price level by cutting taxes even more, but this process would have to stop-once taxes reach zero, they can't be reduced further. We must conclude, then, that high inflation cannot be driven by fiscal policy alone.²

Supply-Side Phenomena by Themselves Produce Inflation?

Because supply shocks and workers' attempts to increase their wages can shift the short-run aggregate supply curve leftward, you might suspect that these supply-side phenomena by themselves could stimulate inflation. Again, we can show that this suspicion is incorrect.

click on figure to enlarge

Suppose that a negative supply shock -- for example, an oil embargo -- raises oil prices (or workers could have successfully pushed up their wages). As displayed in Figure 4, the negative supply shock shifts the short-run aggregate supply curve from AS₁ to AS₂. If the money supply remains unchanged, leaving the aggregate demand curve at AD₁, we move to point 1', where output Y_1' is below the natural rate level and the price level P_1' is higher. The short-run aggregate supply curve will now shift back to AS₁, because unemployment is above the natural rate, and the economy slides down AD₁ from point 1' to point 1. The net result of the supply shock is that we return to full employment at the initial price level, and there is no continuing inflation. Additional negative supply shocks that again shift the short-run aggregate supply curve leftward will lead to the same outcome: The price level will rise temporarily, but inflation will not result. The conclusion that we have reached is the following: Supply-side phenomena cannot be the source of high inflation.³

Summary

Our aggregate demand and supply analysis shows that high inflation can occur only with a high rate of money growth. As long as we recognize that inflation refers to a continuing increase in the price level at a rapid rate, we now see why Milton Friedman was correct when he said that "Inflation is always and everywhere a monetary phenomenon."

---

¹There is a possibility that the short-run aggregate supply curve may immediately shift in toward AS₂, because workers and firms may expect the increase in the money supply, so expected inflation will be higher. In this case, the movement to point 2 will he very rapid, and output need not rise above the natural rate level. (Some support for this scenario from the theory of rational expectations is discussed in Chapter 25.)

²The argument here demonstrates that "animal spirits" also cannot be the source of inflation. Although consumer and business optimism, which stimulates these groups' spending, can produce a one-shot shift in the aggregate demand curve and a temporary inflation, it cannot produce continuing shifts in the aggregate demand curve and inflation in which the price level rises continually. The reasoning is the same as before: Consumers and businesses cannot continue to raise their spending without limit because their spending cannot exceed 100% of GDP.

³Supply-side phenomena that alter the natural rate level of output (and shift the long-run aggregate supply curve at Y_n) can produce a permanent one-shot change in the price level. However, this resulting one-shot change results in only a temporary inflation, not a continuing rise in the price level.

Comments

[Winslow R.]

(high inflation) can occur only if there is a continually growing money supply.

Only?

Wrong.

1) Inflation could also occur with a continually falling GDP and stable money supply.

[Winslow R.]

The problem with this argument is that a continually increasing level of government expen­diture is not a feasible policy. There is a limit on the total amount of possible govern­ment expenditure; the government cannot spend more than 100%, of GDP.

There seems to a bit of confusion between an economy with continually growing nominal GDP and real GDP.

The government can increase expenditure in excess of real GDP forever as long as nominal GDP increases faster than real GDP and never end up spending 100% of real GDP.

Am I missing something here?

[bakho]

Going back to your earlier post, are you suggesting that oil shocks are not inflation and therefore, Volcker was attacking an oil shock and not inflation with his tight monetary policy during the second oil shock in the late 1970s?

If Volcker was attacking an oil shock and not inflation, were double digit interest rates a justifiable monetary policy?

Monetary policy has to operate on real time data. How do you separate the oil shock component from the baseline inflation?

On another note, was Greenspan justified in tightening the money supply in the late 1990s when fiscal policy was extracting 21% of GDP as revenue and spending at a low rate?

[Julio]

The whole article seems to be saying:

There are three possible causes of inflation,
Two of them cannot be sustained for a long period,
therefore the third one (money supply) is the only one that can cause sustained inflation.

Problems I have with this:
1) It assumes those three are then only possible causes of inflation,
2) It redefines short-term inflation as not really inflation. Isn't someone on a fixed income permanently affected by one bout of short-term inflation? Why is this not "inflation"?
3) Can we not have repeated bouts of short-term inflation? Is that not "inflation"?

[James Killus]

Oh jeez, are we back to the monetarist arguments?

Define money supply. Please. Is it M1, M2, M3? Is corporate stock money? How about if it's used as collateral to borrow money? Are government bonds money? How about the U.S. gold reserves? Are dollars used as currency in Albania part of the U.S. money supply?

The money supply is always increasing in a growing economy. If output increases in the right proportion to aggregate demand, there is no inflation. If it doesn't, there is. But output gets measured first in nominal terms, doesn't it? Then one makes an estimate of how the value of money has changed relative to the price of goods. If the supply of goods shrinks, then what? We haven't had such a thing happen in quite a while, but "always and everywhere" is a mighty strong assertion. And "temporary price shock" may wind up looking awfully long term after peak oil happens or if the ocean swallows Florida.

And were the decades long deflations that occurred in farming comunities in the latter 19th century really nothing but a monetary phenomenon? The huge increase in world agricultural productivity had nothing to do with it?

[Outside the Box]

"...inflation refers to a continuing increase in the price level at a rapid rate..."

By that definition, Mishkin is correct. Other factors can increase price levels, but not on a sustained basis. This is consistent with the historical record. For example, before the fed existed, general price levels rose during time of war (fewer consumer items produced), but then fell back to pre war levels after hostilities ended. OTOH, when Rome started adding base metals to their gold coins (thus expanding the money supply), prices rose for centuries.

Of course, other factors can still increase the relative price of goods and services (e.g., medicine) beyond the ability of many people to pay for them. However, these other factors are not under the direct control of the fed, so the fed cannot reasonably be expected to keep them affordable.

The other factors increasing relative price of the selected items (e.g., gov policy) will need to be directly addressed.

An informative excerpt.

[Outside the Box]

The fed is responsible for the dollar losing 95% or so of its value since the fed was chartered.

[jdrietz]

"The money supply is always increasing in a growing economy."

I disagree. The money supply doesn't change by itself - it is caused by human actions. Imagine an economy in which fractional reserve banking is prohibited and only a set number of dollars are available. We would actually be in a continual state of DEFLATION with economic progress.

[Outside the Box]

It seems like this would be the case. However, the charts I have seen of price levels while the US was on the gold standard show price stability, rather than constant deflation. Productivity increased markedly during the period. New gold was mined, but is this entirely the answer? Interesting question to ponder.

[jdrietz]

Perhaps the result of fractional reserve banking?

[Felix]

In a fractional reserve system with an accommodating central bank, Phillips-curve dynamics and the evolution of supply schocks are totally in line with the Friedman statement that inflation is always and everywhere a monetary phenonmenon. Under an endogenous money supply regime, this is insofar true that relative price increases, e.g. due to various price shocks, are followed by money growth which is induced by an accommodating central bank or increased velocity. See for a review: Edward Nelson

[robertdfeinman]

Why is it that almost every graph in an economics textbook looks like the crossed X's in the Exxon logo?

There are always to sets of lines and the statement is made that changing parameter A will shift behavior from one of the X's to the other one and therefore parameter B will change as well. Some how these blindingly obvious (to the author) effects are never backed up with any experimental data. At best we get some appeals to human behavior.

Proposed models are not proof of anything.

[Callahan]

Rising inflation = less money.

I know it is a difficult concept for some.

[reason]

Velocity of circulation anyone? Fix the money supply and eventually someone may find a way to use it more efficiently.

[--Andrew]

Hmm.. If inflation is purely a monetary phenomenon that makes me wonder a bit then about the Fed's current party line concern over inflation.

[James Killus]

If words have no meaning, then "always and everywhere" can mean "sometimes and some places," or even "most times and most places." If words have no meaning.

Alternately, "inflation" can mean "whenever prices rise because of a monetary phenomenon." Then, whenever some non-monetary phenomenon causes a rise in general prices, you can say, that isn't inflation, that's just a general price rise.

Fine. It just brings to mind one of my favorite lines from a sci-fi novel, where the faster-than-light drive was explained as "not increasing your velocity; it just let you go a greater distance in less time."

[Max]

This isn't that hard to understand, guys. It's almost like you are trying not to understand.

BTW, U.S. money supply is going up 11.8% per year. That number minus GDP is a pretty good first cut at inflation.

[reason]

Max...
1. What exactly are we trying not to understand?
2. BTW, U.S. money supply is going up 11.8% per year. - how are you defining money supply?

Are you trying not to understand?

[Winslow R.]

"The false specter of inflation is always raised against such suggestions that our government fulfill its responsibility to furnish the nation’s money supply. But that knee jerk reaction is the result of decades, even centuries of propaganda against government. When one actually examines the monetary record, as The Lost Science of Money does, it becomes clear that government has a superior record issuing and controlling money than the bankers have."

http://www.monetary.org/chicagoplan.html

[Winslow R.]

The young Milton Friedman was the best known advocate for the Chicago Plan in the postwar period, writing: “Henry Simons held the view…which I share - that the creation of fiat currency should be a government monopoly.” Friedman testified on this before Congress as late as 1975 and in 1985 wrote: “I have not given up advocacy of one-hundred percent reserves.” Friedman thought the transition to 100% reserves would not be difficult – “say 25% a year from now, 50% two years from now, etc” (CP 173, 181)

[Winslow R.]

"This is one of the best ways to proceed today – showing the unfairness – that is the immorality - of granting special privileges within our society. Who is going to dare argue for special treatment in principle? For what justification? There is none. Their position is untenable. Their focus on mechanics and ill-defined, and therefore confusing concepts can be VIEWED as a diversion. They are happy to argue over those things forever, so long as they are holding the special privilege – the money power – in the meantime."

[Winslow R.]

I find it is often system instability that brings about the 'required' increases in monetary growth that can result in inflation.

One last quote from Ronnie J. Phillips:

"Establishing separate monetary and
financial service companies enhances the safety of the payments system,
improves the ability of the Fed to control monetary aggregates, reduces government
regulation of banks, accommodates the growth of mutual funds, and
eliminates or significantly reduces deposit insurance."

http://www.levy.org/pubs/ppb/ppb17.pdf

says...

I am struggling with a very simple question but can't seem to find an answer anywhere. What portion of the money supply growth is actually the new (beyond replacement of old bills) currency and coins issued by the Treasury? And... where is the entry point through which this new money gets into the economy? I mean.. noone can get free money, right? Thanks!

[Winslow R.]

"I am struggling with a very simple question but can't seem to find an answer anywhere. What portion of the money supply growth is actually the new (beyond replacement of old bills) currency and coins issued by the Treasury? "

Simple answer fed spends currency (reserves) printed by mint (held at reserve bank) to purchase tsy secs to keep fed funds rate at target.

"And... where is the entry point through which this new money gets into the economy? I mean.. noone can get free money, right? Thanks!"

fed purchases tsy secs from primary dealers

http://en.wikipedia.org/wiki/Open_market_operations

[Winslow R.]

"I mean.. noone can get free money, right? Thanks!"

Actually if you follow it through, the government 'gets' money for free. As the government gets too much money for free we go from an 'inflation' of the money supply to an 'inflation' of the price level.

To avoid raising interest rates thus encouraging savings, the gov could instead increase taxes pulling the 'excess' money from circulation. This is the reason some refer to cash and tsy secs as future tax liabilities.

[I.O.Kitov]

There are alternative opinions on the driving force behind inflation.
Not going to criticize Miskin's view from the point of view of conventional economic analysis - too many researchers made this for me during the last forty years.
The evolution of inflation in the USA after 1960 (data before are not reliable) is governed by only one variable - the change in labor force level. Uncertainty of 2-year ahead GDP deflator prediction using labor force is 0.8 % for the whole period and only 0.4% for the last 20 years.
No model performs better.

http://ikitov.blogspot.com
http://inflationusa.blogspot.com