Back in December, I used one of my weekly Freeman Online columns to address what I saw as a common misunderstanding of how fractional reserve banking works, at least among many who comment on various Internet sites devoted to Austrian economics, especially ones critical of fractional reserve banking. Below, I reprint that column with a few minor changes. Interested readers might also wade through the (70!) comments on the original column if they wish to explore this issue in more detail.
In some free-market circles fractional reserve banking (FRB) is blamed for everything from business cycles to bad breath. Defenders are seen as apologists for inflation and fraud. Thankfully these views remain a minority because they are gravely mistaken. As I, and other Austrian monetary theorists, such as George Selgin and Larry White, have argued, there's nothing wrong with FRB that getting rid of a central bank can't cure. Fractional reserve banking works just fine in a free market.
I don't want to rehearse the whole debate in this column, but I do want to address a claim made by opponents of FRB. They often say something like: "If I deposit $1,000 in my bank and it has to hold only 10 percent reserves, it can create $10,000 in new money." This claim is ambiguous at best and downright wrong at worst. As stated it betrays a lack of understanding how fractional reserve banks (whether under free or central banking) actually work. Let's assume we have a fractional reserve banking system in which banks face a 10 percent reserve requirement. (Note now that we are not talking about a free banking system – I want to make a point about fractional reserve systems in general and show how the problem is that the system isn't free, not that it's based on fractional reserves.)
First of all, this claim is ambiguous about where the deposit comes from and what it consists of. For example, if I deposit a $1,000 check in my bank that you've written on your bank, what happens? It's true that my bank gets $1,000 in new reserves, but it cannot create $10,000 in new loans with the money. Why not? Imagine it credited $10,000 to the borrowers' accounts. What would they then do? They would spend it because that's why people borrow money! And what happens when it's spent? The banks in which the funds are eventually deposited ask the original bank for $10,000 in reserves.
The problem is that if the bank was at its 10 percent requirement before the $1,000 deposit came in, it cannot lose $10,000 in reserves without falling below its minimum requirement (or its desired level, in a free-banking system with no such requirement, which would be unacceptably risky without deposit insurance). What can the original bank afford to lose? Well, it has my new deposit of $1,000 against which it has to keep 10 percent, or $100. Therefore it has $900 to loan out. And that's all. As I call it when I teach "Money and Banking," this is Banking Rule #1: No individual bank can lend more than its excess reserves, in this case $900.
Now you say, "Yes, but that $900 will be spent and deposited at another bank, which will keep $90 and lend out $810, and so on." And you are quite right, which leads us to Banking Rule #2: The banking system can expand by a multiple of those original excess reserves. Assuming 1) all banks face a 10 percent requirement, 2) no one takes wants outside money, and 3) no banks hold excess reserves, the system will create $10,000 based on that original $1,000 deposit. So perhaps the problem with the original statement is that it focused on one bank only rather than the banking system as a whole.
But this is hardly the whole story — and we need the help of our old friend Monsieur Bastiat to see the unseen. If the $1,000 I deposited came from your bank, it loses the $1,000 in reserves transferred to my bank. That forces your bank to call in loans to make up the lost reserves, which leads to reserves being lost by other banks, which then have to do the same thing. The result is that the $10,000 created by my bank's gain in reserves is canceled by the $10,000 destroyed by your bank losing those reserves. When you write a check to me and I deposit it, there is no bank multiplier on net (assuming the three conditions above hold). Thus we see the reverse of Banking Rule #2, as the system simultaneously contracts by a multiplied amount of the original deposit/withdrawal.
So how does new money ever get created and multiplied on net? By injections of new reserves. Only one entity can create new reserves on net in a fiat money system with a central bank: the central bank. When the Fed conducts open-market operations it adds new net reserves to the system, which enables the money-multiplier process with no offsetting loss in reserves elsewhere. The central bank and only the central bank can do this.
A clever fellow might now say, "Well, what if I deposit currency into my bank? There's no offset then, right?" That is indeed true. But where did the currency come from? At some point, you or someone else had to withdraw it from the banking system, which caused a multiplied contraction in the total money supply because currency counts as reserves. The two halves of the process are separated in time, unlike with the deposits, but the net effect in the long run is still zero.
Injections of new currency can cause the money-multiplier process, but guess what is the only thing that can create new currency in a system with a monopoly central bank? You got it: the central bank. If you want to know whom to blame for setting off the money-multiplier process, you need only look there. The monetary base, which corresponds to the total level of potential bank reserves (being the sum of the total supply of currency plus the the supply of bank deposits at the Fed), is totally under the control of the central bank. No one else can create currency and no one else can create net additions to the total amount of deposits at the Fed.
As Robert Higgs points out in a recent blog post, for increases in the monetary base to become increases in the supply of money, the banks have to cooperate by lending out their excess reserves. Banking Rule #1 does not say that fractional reserve banks must lend out their excess reserves, only that they cannot lend more than their excess reserves. Higgs argued in an earlier post that the reluctance of banks to lend out those excess reserves is what is preventing the remarkable increase in the monetary base since the fall of 2008 from turning into significant inflation. Factors such as the Fed choosing to pay interest on bank reserve deposits, the large cash holdings of big firms, and the persistent regime uncertainty that makes lending/investing seem particularly risky these days can together explain the reluctance of the banks to turn the monetary base into money via the multiplier process. Still, it remains the case that only the central bank is responsible for the expansion of that base, even if the banks balk at lending it.
But what about free banking?
In a free banking system, matters are a little bit different. Two factors can, effectively, change the ability of the banking system to initiate that multiplier process. Changes in the supply of the outside money are one such factor. In a commodity-backed free banking system, an influx of that commodity into the banking system brings in reserves and enables the banking system to expand. On the margin, however, the quantity of new commodity money entering such systems will be small compared to the total supply of the commodity in any given period of time. In practice, this has not posed an inflationary problem for (mostly) free banking systems.
Second, in a free banking system, the reserve ratio is determined by the banks themselves, not by the central bank. The ratio need not be treated as an exogenous variable. Free banks can lower their desired reserve ratios which will enable them to create more liabilities off of a given amount of outside money. And it is here that we move from the mechanics of banking to the thornier theoretical issues. If free banks see an opportunity to safely reduce their reserve ratios to enhance their profitability, it's likely because they have perceived that the demand to hold their liabilities has increased, reducing the demand for their reserves via inter-bank and over-the-counter redemption. With fewer claims being made on their reserves, some of their reserves that were previously "desired reserves" are now seen as "excess reserves," and Banking Rule #1 is in play: these now excess reserves can be lent out in the form of a larger supply of bank liabilities (most likely in the form of new deposits granted to borrowers).
From a monetary-theoretic perspective, if free banks create more liabilities when the demand to hold those liabilities has increased, the results will not be inflationary, rather this warranted increase in the total money supply will prevent a deflationary excess demand for money from setting in. The increased demand to hold the bank's liabilities (i.e., the falling demand for its reserves), is a form of savings that drives down the natural rate of interest. When the free bank responds by lowering the market rate it charges to attract the marginal potential borrower on the demand for loanable funds curve, it is not inflating but maintaining the all-important Wicksellian coordination of the market and natural rates of interest. So even if free banks do start to create more money by lowering their desired reserve ratios, this decision faces the test of profit and loss in the marketplace, which will determine if the entrepreneurial judgment of the bankers is correct.
The bottom line is that it is not fractional reserve banking per se that is the cause of inflationary increases to the money supply due to the money multiplier process but rather the ability of central banks to override market signals, thanks to their monopoly status, and add reserves to the banking system at their discretion and independently of the public's preferences. Again, there's nothing wrong with fractional reserve banking that getting rid of the central bank and other government interventions wouldn't cure.