What's Inflation?
A conflicting definition
Any textbook defines inflation as an ongoing increase in the (general) price level. Put it differently, inflation is when (on average) all prices persistently increase over time. If P denotes the price level, then the inflation rate is the percent change in P if said increase continues on time.
Many interpret the textbook definition as identifying inflation with any change in the price level for whatever reason. What matters is that P moves, not why. Under this interpretation, both an excess of money supply and a real shock would cause inflation.
However, the textbook definition does not talk about any increase in P; it talks about “an ongoing rise in the general level of prices.” How long should P increase before we are ready to call it inflation? One month, two months, three months? This is important because while monetary disequilibrium can produce persistent changes in P, real shocks are unlikely to have that continuous effect.
Let’s compare what in this post I label as the positivist and the hermeneutic interpretations of inflation. The former focuses on measurement and sees inflation as any change in P for “whatever reason.” The positivist focus is on observable and measurable changes in P. The latter focuses on understanding the reason why P is changing. The hermeneutic approach is not about measuring inflation, it is about understanding its cause. Let me use an analogy. The flu is a disease, not the fever it produces. Similarly, the disease is an excess of money supply, and the symptom is an increase in the price level.
Mises offers an interesting take on the issue of what is “inflation.” There is no need to reinvent the analysis if we can use the history of economic thought to our advantage.
Mises’ explanation
Mises offers a good summary of what here I call the hermeneutic approach to inflation. Consider first this passage from Human Action (1949 [1996], p. 442, emphasis added):
The notions of inflation and deflation are not praxeological [scientific] concepts. They were not created by economists, but by the mundane speech of the public and of politicians.
Therefore, “inflation” is not a word with a precise definition as one would expect from a scientific term. The underlying idea in the minds of the public and politicians is that there is no inflation when the purchasing power of money is stable. Mises continues (p. 442, emphasis added):
They implied the popular fallacy that there is such a thing as neutral money or money of stable purchasing power and that sound money should be neutral and stable in purchasing power. From this point of view the term inflation was applied to signify cash-induced changes resulting in a drop in purchasing power, and the term deflation to signify cash-induced changes resulting in a rise in purchasing power.
And, since P is not perfectly stable, the begging question is how large the change in P must be to call it inflation? Again, Mises (1949 [1996], p. 446, italics added):
[Individuals] ignore […] perpetual fluctuations as far as they are only small and inconspicuous, and reserve the use of the terms to big changes in purchasing power. Since the question at what point a change in purchasing power begins to deserve being called big depends on personal relevance judgments, it becomes manifest that inflation and deflation are terms lacking the categorial precision required for praxeological, economic, and catallactic concepts.
As long as inflation is defined as the percent change of P, inflation (and deflation) is an ongoing phenomenon. Only changes in P that are “too big” become an “inflation problem.” But what is “too big” of a change is a subjective and arbitrary decision. The same arbitrariness or “unscientific” problem is present on how many times P should change before we call it inflation.
A scientific approach (but not necessarily positivist) to inflation would (a) not depend on an arbitrary decision of what is a significant and persistent change in P and (b) would not confuse the disease with the symptom. Consider, once again, Mises in The Theory of Money and Credit (1912 [1981], p. 272, emphasis added):
In theoretical investigation there is only one meaning that can rationally be attached to the expression Inflation: an increase in the quantity of money […], that is not offset by a corresponding increase in the need for money […], so that a fall in the objective exchange-value of money must occur. Again, Deflation (or Restriction, or Contraction) signifies: a diminution of the quantity of money […] which is not offset by a corresponding diminution of the demand for money […], so that an increase in the objective exchange-value of money must occur.
In terms of the equation of exchange (MV = Py), inflation is when M increases faster than the fall in V. Inflation is the change in P due to changes in MV, not to changes in y. In this approach, not all changes in P are inflation, only those driven by a monetary disequilibrium. Inflation would be the percent change in P cleaned from change sin y.
The textbook definition of inflation seems to be a positivist take (something that must be measurable) on the hermeneutic approach to inflation. It is much easier (and closer to the “mundane speech”) to measure monthly changes in P than changes in MV (recall that V is not observable) adjusted for changes in real output.
So what?
This discussion is mainly about definitions. What matters most is consistency with whatever definition is adopted (inflation is any change in P for whatever reason, or only those persistent in time -produced by monetary disequilibria).
Yet, there is a conceptual difference to keep in mind. If P moves because of an excess of money supply, then the price level is moving out of equilibrium. However, if P moves because of a change in y, the price level is moving towards a new equilibrium. Defining inflation as any change in P carries the complication of distinguishing between “benign (good)” and “malign (bad)” changes in the price level (see Less than Zero by George Selgin).
If you look at the latest CPI report, you will see that the Bureau of Labor Statistics (BLS) carefully talks about increases in the CPI instead of inflation (or the CPI going “up”). The word “inflation” appears only once in a technical appendix. Also, the word “inflation” does not show up on the CPI’s homepage.
There are different ways to define P (CPI, core-CPI, PCEPI, GDP deflator, and so on). But, the different ways to measure P is not the topic of this post.