Is Expanding the Money Supply Costless for Free Banks?

Steven Horwitz

Over at the Mises blog, it is claimed that there is "no cost of production" associated with expanding the supply of money under free banking.  Specifically:

"the bankers can *unilaterlly* affect the demand for their products by simply raising or lowering the interest on loans they offer; they effectively face *no* costs of production. These market “signals” aren’t really market-based at all because their origin in based not on private property but entirely on the will of the bankers. This is what distinguishes them (banks under free banking) from other businesses in the economy as a whole, who face costs of production they cannot unilaterally control."

The bundle of bad economics and misunderstanding of free banking in that quote is quite a sight to behold.

Let's start with the "no costs of production" argument.  It's amazing that this argument even gets raised given that White and Selgin's theoretical models of the profit-maximizing decision facing free banks were all about demonstrating the rising marginal costs facing free banks when they expand.  You can't make their claim that the money supply is determinate without it. 

The key cost is the increased risk of illiquidity as excess supplies of bank liabilities are returned to the overissuing banking for redemption in one form or another.  As free banks expand, more of their notes and deposits will get returned for outside money directly over the counter or indirectly through the bank clearing mechanism.  For a given quantity of outside money (e.g. gold), expanding their liaibilities means a rising risk of being caught with insufficient outside money to pay off those redemptions. This increased risk is not a "physical" cost of production, but it is a cost of production nonetheless. 

Critics need only get as far as p. 4 of White's Free Banking in Britain to see the variable "L (expected liquidity costs)" in his model, along with claim that "expected liquidity costs increase with an increase in [notes] or [deposits]" for a given quanity of specie on the next page.  The bottom of p. 5 and the top of p. 6 catalog the various rising marginal costs facing banks expanding their note issuance and deposit creation. 

So to claim that free banks face "no costs of production" requires either that one is ignorant of the core claim of the free banking literature and/or that one is adopting a notion of "costs of production" that is deeply at odds with the subjectivism and marginalism of Austrian economics.  In either case, this anti-FB argument fails and fails miserably.

The liquidity cost point provides the response to the second point, which is that somehow free bankers don't face costs of production they cannot unilaterally control.  When one understands that the relevant cost of production is the risk of illiquidity, which itself depends on the preferences of liability holders, then there's absolutely nothing different about free banks from any other firm on the market facing an upward sloping marginal cost of production curve.  They are indeed "based on private property" - namely, the private property of the owners of bank notes and deposits who will decide whether or not to exchange that private property for outside money or a deposit credit at another bank. 

It's worth noting that free banks face rising costs of underproduction as well.  To the degree that free banks produce too few notes and deposits, they will face rising opportunity costs in terms of the foregone interest on the loans they could make to bring those notes and deposits into existence.  Finding the joint minimum of the costs of over- and under-issuance is one way to look at the entrepreneurial task facing free banks.

The process by which free banks face rising costs of production and are thereby constrained to produce the quantity of their liabilities their customers wish to hold (and no more or no less) is thus absolutely no different from the process facing every firm in all competitive markets.  That, in fact, is the core of the argument for free banking, which is to apply the (Austrian) theory of the market process to the production of money.  The claim that free banks are somehow different is without merit.

Comments

[Abhinandan Mallick]

Professor Horwitz,

This is my first comment here, and I've avoided entering the controversy on this issue, and I must say arguments on this debate from both sides at times compelling and at times puzzling.

However, the above post is probably the most concise and helpful condensation of the mechanism of free banking I have seen(indeed you've highlighted for me the very reason why FRB is so corrosive in the politicized and cartellised banking system we currently have).

Thanks, from a physics student with an insatiable apetite for praxeological economics.

[Abhinandan Mallick]

Oops, grammatical error, I meant to say "I have found" at times compelling and puzzling.

[Jeff Oxman]

Prof. Horwitz,

I agree with Mr. Mallick. Your illustration of the costs of production is brief, pointed, and clear. Thank you for posting it.

Are you aware of any discussion regarding free banking versus central banking and its relationship to corporate finance? In my readings on the history of corporate finance, it seems that sovereigns typically used some form of imposition on merchants to extend financing to the crown. I wonder if the early encouragement of central banking even by merchants was a way to get around this crown imposition into their business.

[Jerry O'Driscoll]

The argument might be strengthened by considering modern banking practices. Banks purchase deposits to fund the expansion of assets. As these purchases increase, the interest cost rises. As they expand assets, their return on assets fall. They are caught in an interest-rate squeeze apart from the loss of reserves.

[Mike Sproul]

Excellent points Steven. Their error begins with their belief in fiat money. Once they believe that paper dollars are unbacked, and that they have value only because their supply is limited while demand is positive, it's a logical next step to think that money-issuing private banks will reduce the demand for the fiat money and thus cause inflation, as if private banks were no different from counterfeiters.

[Bill Woolsey]

I second Jerry O'Driscoll's point.

I really think there is too much emphasis on zero-interest currency.

White, especially, but Selgin as well, focused on historical episodes where banknotes were the dominant form of bank-issued money.

And, of course, it is possible that there could be banks that specialize in issuing currency even now, or at some future time.

And the argument that applies in such situations is certainly important for theoretical completeness. And, of course, privatized issue of hand-to-hand currency is controversial, and requires special treatment.

But in the real world banks pay interest on the deposits they use to fund loans.

By the MZM measure of the money supply, something like 90% of money takes to the form of deposits. And last I read, something like 90% of payments by dollar volume are by check or electronic equivalent, transfers of ownership of deposits.

If the hand to hand currency were private, then the issue of currency by each bank and the banking system would be limited to the proportion of the total quantity of money preferred by those using money. Any excess currency would be rapidly deposited in interest bearing accounts. And private bank notes are not base money. While a bank depositing currency has more reserves, the bank whose currency is deposited has less reserves.

As Steve understands (and White and Selgin have both explained,) banks can very cheaply issue banknotes to make loans. Deposits are cheaper still. Just some key strokes on the computer, and the borrower has a deposit. But the banks actually making these decisions realize that the borrowers will spend the money and the sellers will deposit the checks in their own banks. Banknotes or checks or electronic payments must be paid by the bank that make the loans.

To fund its loans, a bank must find people willing to hold its currency or deposits. I don't doubt that banks would do things to try to persuade people to hold more of their particular brand of banknotes, and some those things may increase the demand to hold banknotes. But still, the key way that banks are going to get people to hold more deposits is to pay more interest for holding the 90% of the quantity of money on which interest is easily paid.

Banks are fundamentally financial intermediaries. They borrow money and then lend it out. They earn interest on the money they lend and pay interest on the money they borrow. There revenue is the difference between the interest rates they earn and pay. Their cost of intermediation involve the actual activity of banking, obtaining and servicing deposits, making and servicing loans, and for deposits that can be used as money, providing those payments services.

Rather than building a theory of "banks of issue," it is much better to see hand-to-hand currency as just one, minor, way that banks can obtain funding. It is difficult to pay interest one it, and people will hold some of it, even though it bears no interest. That doesn't make it the center of the banking business.

[Peter Boettke]

Bill posts the following at his blog: http://monetaryfreedom-billwoolsey.blogspot.com/2010/05/was-mises-wrong.html

Personally, I think Bill makes an important point.

George Selgin often denies he is an "Austrian", I doubt he would ever deny he was doing "good" economics however. I have often chided Selgin for this -- "if it walks like a duck, quacks like a duck, looks like a duck, etc. etc.". But I am coming around to his position more and more.

What really matters for this debate is not who said what when and where, but whether what they said was true or not. Again, personally, I think Mises said more things that were true than any of his contemporaries, but he did not have an exclusive monopoly on truth in economics. And he did say things that were not quite correct, and in other instances he said things were true but could be stated better.

I put my original post up because I was at a Liberty Fund recreation of a 1974 conference on the Basics of Monetary Economics. Jerry O'Driscoll was the discussion leader, and Larry and George were in attendance. I think the conference was very good, and the discussion was productive. The readings came primarily from Human Action, with some supplemental readings from Jesus H. de Soto and F. A. Hayek.

What ultimately matters (besides in exercises in intellectual history, which as a card-carrying member of HES I find very valuable) is the assessment of the substantive claims in economics being made. Do the claims Mises makes about money right or wrong; how about the claims of White and Selgin? And then we can also address the substantive claims about monetary policy -- would the Chicago 100% reserve approach work to solve our problems; Buchanan's constitutionalization of money; Friedman's monetary rule; Hayek's denationalization of money?

There is a debate going on over at the Cobden Center blog laying out various plans to address the monetary chaos and fiscal irresponsibility that has wrecked havoc in the economies of the UK and US (let alone Europe).

I keep hoping this debate will become productive, rather than spinning wheels, and that we will get to a serious discussion of how to engage modern monetary and macroeconomic professional economists, and we can address the policy discourse of our day.

[Pietro M.]

I'm reading again Human Action now. In the chapter on the theory of indirece exchange Mises is clear in defending free banking, because competition checks inflation through the interbank clearing mechanism.

In the chapter on the business cycle, Mises analyzes the effect of money injected as credit on the coordination of the market process.

A Mises didn't often express his ideas in confused terms, I think that the two positions can be put together by assuming that Mises believed that free banking would reduce inflation very effectively, but there is always the possibility that banks overextend their credit operation and cause a boom-bust cycle, even in the free market.

With a central bank, the interbank market is distorted and does no longer work as a check to inflation because outside money can be created ex nihilo, as free banking theorists have often argued. Under free banking, banks would pay the costs of their errors.

What I still don't understand in the free banking position is how a reflux of banknotes yields intertemporal equilibrium in the allocation of saving. I would guess that the reflux of notes toward the banks has little to do with the sustainability of investment plans: the law of reflux only balances demadn and supply of money NOW, not in the future. But the relevant notion of intertemporal equilibrium requires a much stronger condition than short-run equilibrium in cash holding.

I think that's the reason in which in Prof. Horwitz book monetary equilibrium is not considered a sufficient condition, but only a necessary condition, for intertemporal equilibrium.

However, I still don't see this necessity: investors may ask for bank credit and at the same time overstep the savings constraint, they can increase their cash holding expecting future investment outlays, and may start a malinvestment boom which does not yield a reflux of money. The expansion of bank credit in parallel with its demand (caused by prospective higher investments) may fund malinvestment.

On a more general claim, I think that all the free banking literature dismisses injection effects too quickly. However, microfounding ABCT without injection effects appears to me to be difficult.

[Pietro M.]

I'm quite happy that so far in the debate (not only this post) no reference to ideological (juridical, normative) issues has been made. Holier-than-thou Austrian Economics usually upsets me, while it is quite evident that the points of debate are positive, not normative: they are about the relevant weight of injection effects vs nominal rigidities, the role of aggregate measures of the money supply in the process of coordination, etc.

[Current]

I think that monetary equilibrium isn't a sufficient condition for intertemporal equilibrium.

If an episode of ABCT has occurred then the structure of production will be misaligned no matter what the supply of money is or what the interest rate it. If after that ABCT misalignment there is monetary equilibrium then there will still be a recession.

[topills.com review]

That has not been stable for the last several decades - it has steadily declined and it has recently gone negative.