by Andreas Hoffmann
George Selgin has a much-discussed post over at Alt-M. I agree with most of it. However, I am puzzled by the following statement:
Austrian accounts of the money-creation process often exaggerate the ability of fractional reserve banks to create money “out of thin air,” even while sticking to a fixed reserve ratio, by looking at only one part of the bank money creation process.
[…]
Actually, it isn’t, for the simple reason that, more often than not, a deposit made at one bank involves a corresponding withdrawal of funds from another bank, as when the deposited sum takes the form of a check.
Now, “exaggerate” leaves room for interpretation. And obviously, deposits are often transferred from one bank to another. But the endogenous money view cannot be so easily dismissed.
Ever since Nicholas Kaldor, Post-Keynesians emphasize the endogenous nature of money creation. Recently, the Bank of England, BIS economists and even New Keynesians like David Romer (by replacing the vertical-sloping LM curve with a Taylor rule, and acknowledging the importance of the interest rate target) suggest that money creation is (mostly) endogenous.
In contrast to the exogenous money view you still find in most intro textbooks, endogenous money creation is typically described as follows:
A bank grants a loan to a customer. The bank balance sheet lengthens. There is an extra asset (the bank loan) and an extra bank liability (the deposit at the bank). When the debtor withdraws the deposits for use, the bank is in need of additional financing. The bank might find financing from other banks. If not, there is always the central bank. To prevent interest rates from rising, the central bank will accommodate the needs of the bank. There is no need to attract a deposit before granting a loan!
While Post-Keynesians tend to suggest that this process is fully endogenous, central banks, Romer et al., and some Austrians view the amount of money created as somehow constrained by monetary policy. I’d say the latter is probably more correct as central banks may increase rates to stem a rise in money growth. BIS economists recommend leaning against the wind policies. During crisis periods, however, central banks really accommodate any rise in money demand. The Post-Keynesian view is closer to reality then.
Now, the issue is to know what to make of it. The “Vollgeld” guys (as I see them) are suspicious of banks (typically left-wing). They want to put money creation exclusively in the hands of central banks. The Post-Keynesians tend to prefer regulation to limit lending. The BIS believes in tightening regulation as well. I am not sure what the Austrian view would be in this regard.
I like Hayek’s view:
“So long as we make use of bank credit as a means of furthering economic development we shall have to put up with the resulting trade cycles. They are, in a sense, the price we pay for a speed of development exceeding that which people would voluntarily make possible through their savings, and which therefore has to be extorted from them. And even if it is a mistake — as the recurrence of crises would demonstrate —to suppose that we can, in this way, overcome all obstacles standing in the way of progress, it is at least conceivable that the non-economic factors of progress, such as technical and commercial knowledge, are thereby benefited in a way which we should be reluctant to forgo.” See Hayek (MT&TC, pp. 189–190).
ThinkMarkets 
I am puzzled by it too. For my take on money creation under fractional-reserve free banking, see:
https://www.academia.edu/12531394/The_Uneasy_Case_for_Fractional-Reserve_Free_Banking
I think it is clear in Selgin’s post (first of several he indicates) that in reference to the quoted statement, he is referring to a banking system without a central bank. (And, of course, historically the reason bankers advocate in favor of a central bank is to make it easier for them to engage in credit expansion/money creation.)
Dear Jule Herbert,
I don’t think so. This is not about a world without central banks.
“A bank grants a loan to a customer. The bank balance sheet lengthens. There is an extra asset (the bank loan) and an extra bank liability (the deposit at the bank). When the debtor withdraws the deposits for use, the bank is in need of additional financing. The bank might find financing from other banks. If not, there is always the central bank. To prevent interest rates from rising, the central bank will accommodate the needs of the bank. There is no need to attract a deposit before granting a loan!”
This post-Keynesian argument would have merit IF central banks were willing to allow the supply of bank reserves to expand willy-nilly. But most aren’t. Instead they set policy-rate targets tentatively, in light of macroeconomic goals; and that simple fact vitiates the argument.
Take a simple case: a central bank has a stable price level target. It sets its policy rate at a level it considers for the time being to be consistent with meeting that target. It will accommodate reserve demand changes with open market operations to keep the funds rate equal to the target rate *only* so long as the price level doesn’t change. At first, let’s assume, it doesn’t.
Now suppose that, abstracting from other changes, our bank decides to do a little more lending, as in the paragraph copied above. You say that it can always count on central bank accommodation, for such will be needed to keep the funds rate from rising above target. But for the central bank to accommodate in this case would be for it to allow credit expansion to proceed to a point that raises the price level. The post-Keynesian “endogenous money” fallacy consists of treating a central banks *conditional,* intermediate policy rate target as if it were an *unconditional* target, which just isn’t how things usually work. So long as central bank rate target settings are conditional on ultimate macroeconomic targets, and especially on targets for a nominal variables (like the price level or nominal income), the “horizontal” central bank reserve supply schedule isn’t really all that horizontal after all! In the specific example I’ve supplied, the long-run reserve-supply schedule is actually inelastic w.r.t. an arbitrary increase in bank lending, in the sense meaning that a perfectly well-informed central bank will not accommodate the increase. In that case, a combination of rising overnight rates and interbank reserve losses incurred by the bank (or set of banks) engaged in increased lending ultimately compels that bank to revert to its original lending level.
None of this is saying that a central bank *cannot* accommodate inflationary growth in bank lending as you describe. It is possible, of course. But in that case it is a matter of central bank policy, and not something inevitable. Bank lending doesn’t create its own funding, in other words, unless a central bank decides deliberately to see to it that it does. And any central bank that did as much would have to be heedless of the macroeconomic consequences.
I agree that central banks consider how setting the operational target will affect objectives like inflation or employment.
But I believe we got to distinguish between the decision for the operational target and monetary policy implementation.
Most central banks use an interest rate as operational target. Monetary policy is then implemented by aiming to achieve that target.
If the board targets the price, the quantity has got to be endogenous. Central banks, therefore, commit to implement policy by accommodating any demand at the given target rate.
That commitment is a reason why market interest rates move along with the target rate as it is announced.
A good understanding of the Law of Reflux cures most people of this silly focus on exogenous vs. endogenous money.
http://jpkoning.blogspot.com/2014/09/the-law-of-reflux.html
Of course I understand what it means for the CB to set a rate target and maintain it. But treating a tentative and conditional intermediate rate target value as if it were a fixed parameter, rather than a value subject to routine, deliberate adjustment, which is what the Post-Keynesian “endogenous money” argument does, is a mistaken procedure nonetheless. It fobs off as an inescapable truth about how modern monetary systems function what is in fact a very particular central banking scenario–and one that rarely plays out in fact.
The P-K’s also commit the still more primitive error of failing to distinguish “demand for money” with demand for loanable funds. The “endogenous money” scenario they imagine is one consistent with all conceivable patterns of money growth, and not just with money growth consistent with accommodating prior changes in the demand for real money balances.
In short, while I have issues, and plenty of them, with certain versions of Austrian monetary theory, I consider it on the whole more sound (if also more old-fashioned) than “Modern” Monetary Theory.
I did not defend the post-Keynesian view. I mentioned it. I argued against the way you portrayed money creation in your blog post.
None of the above is meant to say that central banks have no influence via interest rates, conventional or unconventional measures.
[…] a previous post, Andreas refers to George Selgin’s recent discussion of the place of fractional reserve […]