What is Inflation? (part 2)

Last August I argued (echoing a piece by Arthur Laffer) that recent US experience does not fit the etymological definition of the word “inflation” because nothing is really inflating. For example, the Fed is not “blowing up the balloon” by creating a lot of new money, so as to reduce the value of old money. (Laffer’s piece showed a chart of the sweep-adjusted monetary base – one measure of the Fed’s money creation – which at the time was clearly flattening out rather than accelerating.)

Recently I have been confronted with the need to deal with the subtleties in this etymological definition. In discussing the Romers’ recent paper on taxes in the comments section of Econbrowser, I asserted that “Romer & Romer also find that tax cuts increase inflation.” I was beset by (apparent) monetarists who suggested that I was “confusing price increases with inflation” and implied that inflation could not be produced merely by tax cuts because tax cuts don’t involve an increase in the money supply.

My response, in essence, was that, when tax cuts are funded by borrowing, they “inflate” the supply of dollar-denominated assets – in the form of bonds rather than money. I argued that, with more dollar-denominated assets around, the value of the dollar declines relative to goods and services, and that this process should be called “inflation.” They argued that bonds don’t inflate today’s dollar because they represent future dollars rather than current dollars. I argued that bonds are in some ways a substitute for money and therefore reduce the value of money just as an increase in nuclear power production would reduce the value of coal. If you pump up the economy by creating money substitutes, that’s just as much “inflation” as if you were to do so by creating money directly.

How do we know that bonds and money are substitutes? Because raising the price of one increases demand for the other. Raising the price of bonds (which is to say, lowering the interest rate) causes people to increase their demand for money. I gave a practical example of how they could be substitutes: checkable bond mutual funds. The availability of bonds enables me to hold my assets in the form of a checkable bond mutual fund, which I can use very much as I would use “money” in the form of bank deposits.

The ultimate issue here is money supply vs. money demand. Increases in the general price level are clearly to be considered inflation if they result from increases in the supply of money. But what if they result from decreases in the demand for money? That’s what happens when the government issues bonds, and that’s presumably the main reason that the Romers find tax cuts to be associated with inflation.

It seems to me that the definition of inflation becomes quite flimsy if you don’t allow it to take into account the demand side of the money market. In principle, the general price level can rise without bound just because of declines in the demand for money. When I was in graduate school, one occasionally heard the phrase “self-generating hyperinflation” when discussing monetary theory. Such a hyperinflation did not necessarily require any increase in the supply of money. It resulted from a loss of confidence in the value of the existing money supply. (After all, money has no fundamental value; it’s value depends entirely on individuals’ confidence that other individuals will consider it valuable.) In the event of such a self-generating hyperinflation, it would be little comfort to hear someone say, “Don’t worry; there has been no increase in the money supply; this is not inflation; it is merely price increases.”

One might still argue that the term “inflation” can’t apply unless something is “inflating.” In the case of a “self-generating hyperinflation,” the increase in the price level results from the spontaneous functioning of the market; there is nobody metaphorically “blowing up the balloon.” In the case of a budget deficit, however, there is somebody blowing up the balloon: the government is blowing bonds into the balloon, and even though their effect is arguably indirect, these bonds have the same effect (not dollar-for-dollar but in general) as blowing money into the balloon. Perhaps one could define inflation as “a rise in the general price level produced by a change in the supply of and/or demand for money induced by public policy or other factors exogenous to the money market.” (It should certainly include, for example, the effect that a new gold discovery would have on an economy that uses gold as money – or the effect of a new silver discovery on an economy that uses both gold and silver for exchange but measures prices in terms of gold.)