I’m inclined to agree with Henderson that Sumner’s point is an important one, but at the same time I feel that Sumner only captures half the story. It is absolutely correct to argue that only money can cause an economy-wide disruption. However, monetary disequilibrium is only one way money can affect the real economy. If we frame the question in terms of MV = PT, I’m afraid that monetary disequilibrium only explains fluctuations in velocity (V). There’s another important monetary variable, which is the money supply itself (M). The issue is that the causes of changes in velocity are easier to observe: an increase in preferred cash balances, for whatever reasons. Causes behind significant fluctuations in the money supply are more indirect, and therefore more difficult to discern.
One of Hayek’s most important justifications for his own theory of business cycles was exactly that only money can systematically disrupt the real economy. But, his theory of the business cycle is one which attempts to explain changes in “M,” rather than in “V.” Thus, an excess supply of money can lead to severe intertemporal discoordination, drawing the means of production towards earlier stages of production. The entire structure of production is altered in a way that makes it more capital intensive, but without the requisite capital to complete these projects. Typically, with an increase in real savings, the cost of capital will fall. During an “artificial” boom, however, the price of capital doesn’t fall. Instead, what makes it temporarily affordable is excess money substitute, entering the economy through the banking system. When credit expansion slows or ceases, the unsustainability of the more capital intensive structure of production is revealed and there is a mass failure of investment.
This is a real shock, but it’s a real shock caused by monetary distortions. The same wouldn’t occur if OPEC had sufficient market power to raise the world price of oil, or even with a incredibly damaging terrorist attack. I agree with Sumner that not even a banking crisis alone can cause the business cycle. Instead, financial crises are symptoms of greater problems. Business cycles are demand side phenomena, which by definition ascribes to it monetary causes — remember, even the Mises–Hayek business cycle theory is a demand-side explanation.
The major difference is that I don’t consider real shocks irrelevant. Money and the real economy go together. Money is non-neutral, meaning that changes in the distribution of nominal income also causes changes the distribution of consumers’ and producers’ goods. The best way to understand this perspective is by looking at the business cycle as a subset of the wider field of economic calculation.
Most discussions of complex economies begin without money. While a complex, moneyless division of labor is extremely unrealistic, it helps us understand the point that there is an optimal distribution of resources. This is where every resource is being employed towards its highest valued end. Once this is understood, we can incorporate money. Money is what helps this distribution of resources become possible. But, money’s purpose is not only to lubricate indirect exchange. It has an equally important, and related, role in making possible the emergence of money prices. In a perfectly competitive market, prices will reflect the value of a good. In reality, prices are imperfect, but competitive pressures will keep these prices moving. This aspect of disequilibrium is what gives us profit and loss, and it is this profit and loss which provides the constraints to resource distribution. Unprofitable distributions will go bankrupt and capital will be re-distributed, and there is always an incentive to exploit areas where the gap between the actual distribution and the optimal distribution is widest (i.e. where profits are highest). All of this is thanks to money.
Unequal changes in income distribution — as they tend to be — will change individuals’ purchasing power unequally, allowing some to bid resources towards them at the expense of others. This can be accomplished by increasing the quantity of money. Increases in the money supply will also increase prices, and if the distribution of excess money is unequal so will the change in prices. Imagine we start at the general equilibrium position, then non-proportional and non-simultaneous changes in prices will affect the distribution of real goods by moving away from equilibrium. In the parlance of economic calculation, uneven increases in the money supply will distort the coordination of resources. This is pretty much Hayekian business cycle theory: increases in the money supply will cause alterations away from equilibrium, requiring a movement back to equilibrium when prices adjust to reflect the true value of non-money goods.
Putting all of this in the context of monetary disequilibrium, I am afraid that Sumner focuses too much on money shortages. The market monetarist explanation of the consequences of excess money is not as developed — terms like “overheating” are unsophisticated and do not provide a lot of insight. This is what, in turn, probably leads too little emphasis on the real economy. In the case of money shortages the real economy simply cannot function well, because indirect exchange is hampered. In this case, it makes sense not to pay attention to real shocks. But, in the case of excess money what happens is the distortion of money prices, and therefore profits, and unsustainable changes in the distribution of real goods. A sudden reversion of profits — hence the term “phantom profits” — is a powerful money shock that creates an equally as powerful real shock.
The difference in direction of approach also explains the differences in policy recommendations. To Sumner, a fall in NGDP should be met with increasing the money supply. But, this is because to him the business cycle is basically a change in the velocity of money. If you look at the business cycle from the opposite angle, however, the causes of the reduction in output are much different. Instead, firms declare bankruptcy and excess money is eliminated through loan default and a temporary drying up of credit channels. Increases in velocity can occur, but this is a secondary consequence, probably as a result of heightened uncertainty. And, in this case, an increase in velocity adds to the stock of real savings, making the real adjustment much less painful than otherwise necessary.
Money does matter, but looking at money alone is the wrong way to go about business cycle research. Money is here to help develop a complicated division of labor and coordinate the distribution of resources within this economy. Monetary distortions cause real distortions, and this is a story that looking at changes in money alone can hide. There’s a lot of Austrian business cycle theory in this post, but my point makes just as much sense without it. I just know the theory well, and so I use it as an example. It could be that monetary distortions change the real economy in a completely different way. But the implications are at least similar.