The Form of Saving Matters

Although Austrian economists agree on most of the "big picture" issues, even so there are some internal controversies. One of the most heated debates within the Austrian camp centers on the economic benefits (or lack thereof) of the practice of fractional-reserve banking. Just about every Austrian today thinks that our current financial system — cartelized and propped up by the government — is prone to excessive inflation and the boom-bust cycle.

Even so, many prominent Austrian banking experts believe that in a truly free market, private institutions could usefully engage in limited fractional-reserve banking (see, e.g., Steve Horwitz's book). On the other hand, Murray Rothbard and his followers thought that fractional-reserve banking per se was illegitimate and economically harmful.

In the present article I respond to a very thoughtful analysis from Bill Woolsey, who tries to show the economic equivalence between conventional household saving through bond purchases and household saving via the accumulation of more money created by the banking system. I'll point out that there are important differences between the two scenarios, and that Woolsey has not really met the Rothbardian challenge to fractional-reserve banking.

Woolsey's Case for Bank Money Creation

In the interest of brevity, I can't reproduce Woolsey's full blog post, but it really is one of the clearest and most succinct expositions of what is dubbed the "free-banking" position on these matters. Below I excerpt just enough to bring the reader up to speed:

My view is that only an excess supply of money can possibly lead to malinvestment. If the quantity of money and supply of credit rises to match an increase in the demand to hold money, the consequences are the same as an increase in the supply of credit due to an increase in the demand for bonds.

Let's stop right there. The standard Misesian/Hayekian explanation of the business cycle is that the commercial banks arbitrarily increase the supply of loanable funds, even though the community hasn't actually increased its real savings. (I describe how a commercial bank can "create money out of thin air" here.) In order to lend out the newly created funds, the banks lower the actual market rate of interest below the "natural" rate. At the false, artificially depressed interest rate, businesses borrow too much money and begin long-term projects that society doesn't have enough actual savings to complete. This is what happens during an unsustainable boom, such as the recent housing bubble.

Just about all modern Austrians agree with this basic story. The disagreement arises in the case when the commercial banks increase the supply of loanable funds in response to a sudden increase in demand to hold money. In this case, people like Steve Horwitz and Bill Woolsey believe that the banks do not set in motion a boom–bust cycle. On the contrary, they think that the banks merely offset painful readjustments that would otherwise have to take place to restore equilibrium between the supply and demand for actual money balances. In this scenario, claim the "free bankers," the banks' creation of new money doesn't disturb the economy at all, because in the new equilibrium, producers haven't been given an artificial incentive to start projects for which there is inadequate real savings.

Let's return to Woolsey and let him make his case:

[The free bankers'] approach focuses on a scenario where an increase in the demand to hold bank deposits or banknotes is an increase in saving. Households reduce consumption out of current income and accumulate bank liabilities. The banks expand their lending, issuing new money to borrowers. The monetary equilibrium theorists argue that the increase in the quantity of money just offsets the decrease in velocity, and so there is no addition to the stream of monetary expenditures to cause malinvestment.

The Rothbardians, however, insist that the … [new] money issued to fund bank loans … still distorts relative prices because it doesn't go exactly to those individuals who demanded more money by reducing their monetary expenditures. They accuse the monetary equilibrium theorists of "macroeconomic thinking."

I side with the monetary equilibrium theorists in this debate. The element of truth in the Rothbardian argument is that there is little doubt that the process of money creation does impact relative prices. Where he and his followers go wrong is in assuming that any changes in the pattern of relative prices resulting from an increase in the quantity of money that matches an increase in the demand to hold money is usefully characterized as malinvestment.

The best way to see this is to compare the effects of a change in the demand for bonds with a change in the demand for money. Following the conventional approach, I will start with a scenario where there is an increase in saving by households. Some particular households spend less of their incomes on consumer goods and services. Those households save. Savings in aggregate increase as well. This saving could take the form of purchasing bonds or it could take the form of accumulating money.

First, suppose that some households increase saving, reducing their consumption out of income and instead use the money they earn to purchase corporate bonds.…

Woolsey then spells out how the economy readjusts to the higher level of saving — in the form of more corporate bonds held by the public — by sustainably lengthening the structure of production.

[Now, on the other hand, s]uppose that the households increase saving by reducing consumption expenditures out of income, exactly as before, but instead of spending the money on corporate bonds, they instead increase their money balances. The effects on those from whom the households would have purchased the consumer goods are the same. They sell less and earn less. They hire less labor and use fewer other resources. The incomes earned by those workers and other resource owners fall.

Assume that the banks expand their lending at this same time, and lend out newly created money exactly equal to the amount of money the households have accumulated. The quantity of money expands exactly the same amount as the demand to hold money.

Further assume that the banks expand lending by purchasing corporate bonds, the exact same corporate bonds that the households purchased in the previous scenario. The market interest rate decreases exactly in the same way as before. The added demand for corporate bonds raises their prices and lowers their yields. This creates the exact same signal and incentive for firms to purchase additional capital goods. As before, the firms purchasing the capital goods finance the purchase by issuing new corporate bonds.

Those selling those same capital goods receive greater profits. The increased profitability of producing additional capital goods creates a signal and incentive, exactly as before, to direct more financing into this field of endeavor, and so raises the demand for labor and existing capital goods to expand the production of capital goods.

Using Austrian terminology, the change in demands, away from consumer goods and to capital goods, and the decrease in the market interest rate, results in a shift in the structure of production away from lower order goods to higher order goods. Less production is directed towards the production of consumer goods in the near future and more towards the production of consumer goods in the more distant future. Again, the gap between the prices of resources and their products shifts to reflect the new lower interest rate.

Let's review to make sure we understand Woolsey's point: He first walked us through a standard scenario where households save more by cutting back on consumption spending, and by increasing their purchases of corporate bonds. (In this scenario, the banks don't mess with the total supply of money in the economy.) The switch in household spending causes consumer sectors to contract, interest rates to fall, and higher-level capital-goods sectors to expand. This is the Austrian description of normal, sustainable economic growth through savings and investment.

Woolsey then argues that the same physical adjustments would occur in the structure of the economy if the households decided to increase their saving not by buying corporate bonds, but instead by accumulating larger cash balances. Now, if the banks stood idly by while this process unfolded, the prices of goods and services (in general) would fall, because households want to hold more money, but there's the same amount of money to go around.

However, if the banks intervene and create more money — in the exact quantity that households want to increase their cash balances — then there is no reason for the overall level of prices to change. Some prices still go down (in the consumer-goods sectors), while others go up (in the capital-goods sectors). If we further suppose that the commercial banks "create" the new money by granting new loans to businesses in the capital-goods sectors, Woolsey concludes that the economic transformation is identical to the standard case. So how could Rothbardians endorse the former scenario but reject the latter?

Two Problems With Woolsey's Analysis

In an online article, I obviously can't give a definitive treatment to these complex matters. But I want to point out two major shortcomings in Woolsey's analysis.

First, Woolsey is simply begging the question when he claims that the money created in the second scenario would flow precisely to those businesses that would have benefited from household bond purchases in the first scenario. To his credit, Woolsey acknowledges that his story relies on "heroic assumptions." But he doesn't seem to realize that one of his heroic assumptions is, "The Rothbardians are wrong when they accuse the free bankers of engaging in crude 'macroeconomic thinking.'" It is not surprising, then, that Woolsey is able to conclude that the Rothbardians were wrong to accuse their opponents of relying on crude macroeconomic thinking.

This is a crucial point, so let me restate it: With a fixed stock of money in the economy, suppose some households decide to save more by restricting their purchases of consumer goods (TVs, iPhones, etc.) and accumulating extra cash balances. At the same time, the other households in the community want to maintain their original levels of saving, investment, and cash balances.

Overall, the community's demand to hold money has increased, while the stock or supply of money is constant. This means that the "price" of money must increase, which is the same thing as saying that its purchasing power must increase. In other words, prices of goods and services in general must fall.

In the new equilibrium, the households who want to accumulate larger cash balances will have more dollar bills in their bank accounts and/or in their wallets, whereas the other households will have smaller bank accounts and currency holdings. However, because prices in general would have fallen, the households who saw their money holdings drop will still have the same purchasing power in their (smaller) cash holdings.

On the other hand, the households who wanted to increase their money holdings would actually have more purchasing power than one would think by just looking at the increase in the number of dollar bills to their names. For not only have they taken possession of a larger fraction of the (fixed) number of dollar bills in the economy (both in the form of checking account balances and actual currency), but now prices in general are lower. So the amount of purchasing power in their overall wealth has increased, which is just what they had intended when they began accumulating money.

Here's a tricky point: The change in some households' behavior won't merely rearrange the community's holdings of cash balances and lower the price level. It will also lead to a lengthening in the structure of production. This is because the households who accumulated extra cash didn't cut back on their expenditures symmetrically, but instead focused their cutbacks on consumer goods. In other words, the households who wanted to hold more money didn't say, "This month we're contributing less to our Individual Retirement Account and less to the local diner." On the contrary, they said, "This month we're still investing the same amounts from our paycheck in life-insurance premiums and our IRA, but we're cutting back on dining out, and then socking that money into our checking account."

In the grand scheme, what happens is that the "general price level" falls because the prices in consumer sectors fall, while prices in capital goods sectors stay about the same. This leads to a shift of actual, physical resources away from making TVs and iPhones, and into the production of drill presses and other tools.

We are finally ready to reiterate the Rothbardian objection to the so-called free-banking argument: If the commercial banks see that the general price level is falling because some households are trying to accumulate larger cash holdings, and the banks respond by creating more money, then this process in practice cannot possibly reproduce the same sectoral adjustments, with the only difference being that the overall price level will stay constant.

There are many reasons for this. In the first place, suppose that the banks somehow could isolate the households who wanted to accumulate more cash, and simply gave it to them. This action would eliminate the need for prices in general to fall. But at the same time, it would also obliterate the differential spending on consumer- versus capital-goods sectors. Before, when a household wanted to, say, build up an extra $500 in its checking account, the household would have to cut back on $500 worth of spending on restaurants, movie tickets, etc. But now, if the banks simply add $500 to certain households' checking account balances — without subtracting numbers from anybody else's checking accounts — then these households will have achieved their objective without needing to cut back on spending anywhere else.

Of course, in practice the banks don't "offset an increased demand to hold cash" in this fashion. But I walked through that point to show that if they did, then clearly they would not be "restoring equilibrium" but would instead distort the saving "signal" that the households would have sent.

So let us suppose that we still have households trying to accumulate more money balances, which is starting to cause a general fall in prices. This time, the banks intervene by creating more money. But rather than simply handing it over as a gift to those households who wanted to save more, instead the banks offer loans to businesses at a lower interest rate. This expansion in the amount of loans creates more total money (in the form of checking account reserves) in the community, so that now the households can increase their money balances without causing a general decline in prices.

Here is the crucial Rothbardian objection: What reason do we have to suppose that the people who get the new bank loans in this scenario will be precisely the ones who would have sold more bonds in the scenario where the household savers simply bought more bonds? In general, the Rothbardians argue, the banks' creation of new money will benefit the clients of the expanding banks, endowing the bankers and their prime clients with extra purchasing power, at the expense of everyone else in the community. By focusing on broader aggregates such as "supply and demand for money," the free bankers are overlooking these issues of who gets the new funding, under the two scenarios of (1) households buying more bonds and (2) private banks issuing new loans.

To repeat, Woolsey thinks he has answered this objection by telling a story in which the banks do lend money to precisely those firms who would have sold more corporate bonds to the public, in the first scenario. And then Woolsey admits that this is a "heroic assumption." Does everyone see the problem here? The Rothbardians level an objection, saying the free bankers are ignoring an important real-world consideration, then Woolsey assumes away the problem and declares that he has met the Rothbardian objection.

The Form of Saving Matters

There is a second problem with Woolsey's entire discussion: He acts as if the form of saving were irrelevant. But in the real world, surely it matters to the individual household whether additional savings take the form of a higher cash balance or a larger stockpile of corporate bonds. The obvious difference is that cash balances don't earn interest but they are much more liquid than corporate bonds.

Therefore, I challenge the entire premise of Woolsey's argument. He is trying to show that the injection of new money by banks, can make increased household saving of money the same as increased household purchases of bonds. But that can't possibly be right, because in the real world it definitely does make a difference if households want to collectively stockpile more cash rather than stockpiling more corporate bonds.

Because Woolsey's simple model can't account for the distinction between savings in the form of money balances and in the form of bonds, we should be suspicious of its use to validate the free-banking position on a question where this distinction is important.

Conclusion

Bill Woolsey has offered a very clear and succinct summary of the case for free-market fractional-reserve banking. Yet he hasn't really addressed the Rothbardian objections. Instead he has assumed them away by creating a model where the objections aren't allowed to operate.