A Note on Some Recent Misinterpretations of the Cantillon Effect
Abstract: Book (2019) claims that Austrian economists attach too much weight to Cantillon’s discussion of monetary redistribution, while Sumner (2012a, 2012b, 2012c) argues that it makes very little difference how new money is injected. In this note, I critically review these arguments, finding that they are unconvincing. The Cantillon effect matters, and the Austrians correctly analyze it.
inflation — central bank monetary policy — austrian economics — cantillon effect JEL Classification: B11, B53, E31, E51
Arkadiusz Sieroń (firstname.lastname@example.org) is assistant professor of economics at the Institute of Economic Sciences at the University of Wroclaw, Poland.
It may seem that the Cantillon effect is an easy-to-grasp and noncontroversial concept. After all, the observation that helicopters do not drop money and that money goes into the economy in an uneven manner—that is, some people get new money earlier or get more of it—should not elicit doubts. However, almost two hundred years since its formulation, the Cantillon effect still causes a lot of confusion. The most recent example is Joakim Book’s article “The Mythology of Cantillon Effects” (2019), in which he argues that the weight some Austrian economists “attach to Cantillon’s monetary redistribution is greatly exaggerated.” Another case is Scott Sumner’s series of blog posts (2012a, 2012b, 2012c) in which he claims that it does not matter how money is injected into the economy. The aim of this brief note is to critically review Book’s article and Sumner’s posts.
II. BOOK’S MYTHOLOGY OF CANTILLON EFFECTS
Book acknowledges that “by introducing new money into the system, early receivers are benefited at the expense of late receivers.” However, he points out that Cantillon’s analysis of monetary redistribution is based not on central bank fiat money, but on a gold standard: “For Austrians, usually strong proponents of hard money, commodity-money regimes, and monetary policy rules, it is ironic that the monetary-redistribution analysis that so endeared him to Austrians is based entirely on a gold standard.”
Hence, the alleged problem with the Austrian view is that it incorrectly considers the Cantillon effect “a fundamental evil unique to fiat central bank systems,” while in reality “every monetary system includes Cantillon effects.” Book has it right that “Cantillon effects are universally valid occurrences of any monetary economy.” But his claim that Austrians are not aware of this fact is false. Surely, the first-round effect is a result of any uneven increase in money supply, but that does not mean that the Cantillon effect is irrelevant or that various possible ways in which the money supply in a given economy can be increased have the same results.
On the contrary, the great merit of the Austrian school is to notice that there are differences in the outcomes of monetary inflation, depending on its variant—in particular, whether new money is introduced into the economy through market or nonmarket channels. As I point out in my book Money, Inflation and Business Cycles: The Cantillon Effect and the Economy, in the former case,
money supply increases as a result of voluntary activities of market participants involved in the production of money (or its transfer from abroad) under private ownership. Market production of money is undertaken for profit, which can take place on a completely unhampered market only by adequately meeting the needs of the money users, as a result of providing an appropriate amount of universal medium of exchange of appropriate quality. This means that such production has a balancing effect. If the increase in money supply is too big, prices will rise to the point where the purchasing power of the money produced will drop to the level at which further production will no longer be profitable. Market production of money is therefore subject to a self-regulating profit-and-loss mechanism. In contrast, the creation of money can be practically unlimited. While in the case of fiduciary media, there are some physical or institutional limits resulting from the limited amount of money proper, on the basis of which they are issued, in the case of fiat money, there are no such restrictions. Its supply can be increased until the monetary system collapses. (Sieroń 2019, 66)
Book’s second charge is that “Austrians also overstate their case, at least as far as Cantillon’s writing is concerned; to our ‘man of mystery’ monetary redistribution does not inevitably set the economy on a path of unsustainable boom and bust—relative prices, wealth, and consumption desires adjust.”
Book has it right again. But nobody claims that each increase in the money supply and its related Cantillon effects trigger the business cycle. Yes, the Austrian business cycle theory says that the creation of money is responsible for the business cycle, but only if the newly created money is introduced into the economy through the credit market. We could even say that that theory is essentially an analysis of one particular variant of the Cantillon effect, as “it examines how the increase in money supply by a particular channel—the credit channel—affects the specific price: the interest rate, leading to changes in the structure of production” (Sieroń 2019, 46).
The last point I would like to address is Book’s claim that “to Cantillon all new money had the same redistributive and uneven effects, regardless of whether it was first spent in the real economy or entered the credit markets, reducing interest rates.” This statement is blatantly false. In general, Cantillon was perfectly aware that what is important for the economy is not only the fact that money supply increases, but also the way it happens. He stated:
The proportion of the dearness which the increased quantity of money brings about in the State will depend on the turn which this money will impart to consumption and circulation. Through whatever hands the money which is introduced may pass it will naturally increase the consumption; but this consumption will be more or less great according to circumstances. It will be directed more or less to certain kinds of products or merchandise according to the idea of those who acquire the money. Market prices will rise more for certain things than for others however abundant the money may be. (Cantillon 1959 , II.VII.6)
In particular, Cantillon acknowledged that the increase in the money supply does not always lower interest rates, because it depends on how the monetary inflation occurs. He wrote:
If the abundance of money in the State comes from the hands of money-lenders it will doubtless bring down the current rate of interest by increasing the number of money-lenders: but if it comes from the intervention of spenders it will have just the opposite effect and will raise the rate of interest by increasing the number of Undertakers who will have employment from this increased expense, and will need to borrow to equip their business in all classes of interest.
Hence, if Cantillon understood that the increase in the money supply lowers interest rates in some cases while raising them in others, he could not have believed that new money has the same redistributive and uneven effects in both cases.
III. SUMNER’S CLAIM THAT IT DOES NOT MATTER HOW MONEY IS INJECTED
Book is not the only economist who has recently misinterpreted the Cantillon effect. Another is Scott Sumner, who, in a series of blog posts (2012a, 2012b, 2012c), argues that it makes very little difference how new money is injected, or, to be more precise, who gets the new money first. He supports his thesis with three arguments: (1) the new money has no more purchasing power than the existing money; (2) the new money is sold at fair market prices; (3) when the central bank purchases assets, the money is just swapped for securities, so the recipient of new money is no wealthier than before the transaction.
Let us analyze these arguments. First, it is true that each monetary unit is the same and one cannot distinguish the new and old money. However, the point is that price inflation is a sequential process. Hence people whose cash balances increase before the prices adjust to the monetary injection have more purchasing power than people whose cash balances increase after the prices adjust, even if all monetary units have the same purchasing power. Or, the point is that the monetary injection dilutes the purchasing power of the whole money stock, even if all monetary units have the same purchasing power.
Second, it is true that the central banks do not give away money for free, but purchase assets in the secondary market, swapping money for securities. But Sumner misses the point. The crux of the problem is that when the central bank creates money, it, just like a money counterfeiter, enters the market with some unearned purchasing power. This enables the central bank to possess more assets and to exert greater influence on their prices, affecting not only the price level, but also the structure of prices. Sumner suggests that for the sellers of assets, the central bank purchases are just one of many transactions, so they are no wealthier than before. Abstracting from the fact that individuals benefit from exchanges, the point is that the central bank’s transactions are special because they introduce the new money into the economy and thus are inflationary. The whole idea of the Cantillon effect is that during monetary inflation, it’s better to spend money before rather than after price adjustment. Those people who get money first rather than those people who get money last have simply more chances to do it.
Although the Cantillon effect describes the simple idea that new money enters the economy at specific points, it still causes some confusion. Two recent examples are Book’s and Sumner’s objections. Both critiques are invalid.
Book correctly notes that the Cantillon effect occurs under every monetary regime, but that does not negate the particularly negative effects attributed by the Austrians to the increase in the supply of fiat money under central banking. This realization of the Cantillon effect, especially if new money enters the economy via the credit market, is particularly harmful, which only strengthens the case for the detailed examination of how exactly the new money enters the economy.
Sumner does not claim the Cantillon effect occurs under every monetary system, but he does claim that it does not matter, as the economic effects do not depend on who gets the new money first. He believes so because he treats central banks’ purchases as just swapping one asset for another. However, in reality, central banks increase the money supply through asset purchases that change the cash balances of certain institutions. These sellers value then each monetary unit less and therefore increase their spending, leading to changes in the relative-price structure and production and to fluctuations in the cash balances of the next agents, which in turn leads to repeating the cycle in subsequent rounds (Sieroń 2019). Hence who gets the new money first definitely matters.