6. On the Definition of Money

6. On the Definition of Money

April 1959

The recent excellent articles by Gordon W. McKinley and Donald Shelby highlight the importance of a still unresolved problem: the proper definition of money.⁴³ When McKinley likened the argument that savings deposits (in contrast to demand deposits) are only money “substitutes” to the old doctrine that demand deposits are themselves substitutes rather than true money, he touched briefly on a highly important point in the elusive problem of defining “money.” Nowadays, it is simply and casually accepted by all economists that demand deposits are part—and the largest part—of the money supply. Yet the reasons for this acceptance have been forgotten and, in the process, important insights into money have been foregone.

For it should not be forgotten that demand deposits are only money so long as the bank is considered safe; let the bank be thought in imminent danger of failure, and a bank run develop, and then its demand deposits will no longer be readily accepted at par, and will no longer function as part of the money supply. This truth is obscured nowadays because of the public faith that the FDIC will protect any bank from runs; but it was well known before 1933, when bank runs were often heavy and even endemic. Furthermore, in the pre-Civil War days when all banks printed their own notes, the notes often circulated at great discounts. This was especially true beyond the home areas, where people did not have full confidence in the bank’s ability to redeem.

This dependence of the moneyness of demand deposits on public confidence is not a function of the gold standard; it is true today as well (or would be in the absence of the FDIC). The point is that whether the ultimate standard—the ultimate money—is gold or paper, demand deposits of commercial banks are not that standard; they are only redeemable in standard cash. In the strict and narrow sense, only the ultimate money, that which is not redeemable in something else, is truly money.

To confine the term “money” thus to legal tender would be a cogent definition, but not a very useful one. It would not be useful because such money-substitutes as demand deposits play the role of money in the economic system: they act as would an increase in money in their effects on the economy. They do so precisely because people believe that they do stand for money, that they are redeemable in cash. As long as people continue to believe this, they are willing to exchange demand deposits, or to hold them in their cash balances, as absolutely equivalent to money. Both are equivalent “dollars.” Thus, demand deposits may be considered, so long as they are thus equivalent, as part of “money” in the broader sense.

But what of such assets as savings deposits? Those who would confine money-in-the-broader-sense to demand deposits assert that the two are uniquely different: that the latter are used as media of exchange while savings deposits are not. But this difference, while important in many respects, is not at all decisive here. For here we have only a difference in the form in which money is kept. Suppose, for example, that through some cultural quirk, everyone in the country decided that they would not use their five-dollar bills in exchange. They would only use their ten- and one-dollar bills, and keep their cash balances in fives. As a result, the five-dollar bills would tend to circulate far more slowly than the other bills. Now suppose that, when a man wants to reduce his cash balance, he may not spend his five-dollar bill directly; he goes to a bank and exchanges it for five one-dollar bills, which he proceeds to trade. In this hypothetical situation, the status of the five-dollar bill would be exactly the same as the savings deposit today. The five-dollar bill—like the savings deposit—would never be used as a direct medium of exchange; and, again like the savings deposit, it could only be used as a medium at one remove: the holder must go to a bank and exchange it for that type of money which will serve as a medium. Yet would anyone say that the five-dollar bill is not part of the money supply? But if it is, we must also say that the savings deposit is likewise part of any broader definition of money that includes demand deposits.

In short, a savings deposit is money of a different form than demand deposits. It circulates more slowly, and more of people’s long-term cash balances are held in this form; and yet, whenever the holder wishes to use it as a medium, he simply goes to his bank and obtains the demand deposit. Or, indeed, he may just as readily obtain cash directly, as he could with a demand deposit.

Those who stress the use of demand deposits as direct media forget why they are media in the first place: because they are generally regarded as redeemable at par for cash—the original medium. But then are not any assets, likewise redeemable at par, also part of the money supply? If so, then savings deposits, both in commercial and savings banks, must be considered as part of the money supply because they are redeemable at par. And, furthermore, so must we consider savings-and-loan shares, which the savings-and-loan association promises to redeem at par.⁴⁴ But McKinley and other writers who have argued persuasively for the inclusion of savings deposits and savings-and-loan shares in money have neglected other assets: notably, government savings bonds and the cash-surrender values of life insurance. Government savings bonds, with their fixed guarantee of redeemability in cash, certainly are just as much money as savings deposits. To be precise, of the nonmarketable Treasury liabilities, savings bonds, savings notes, and Series A investment bonds (redeemable in cash whereas Series B bonds are not), must be treated as part of the money supply.⁴⁵ Marketable securities, like all other assets, are exchangeable for money, but only at varying and nonfixed rates, and are therefore not money but “goods.” And this is true even for short-term government bonds, which are highly liquid and can function as “near-money” substitutes for part of a person’s cash balance; for they are only exchangeable for money at the market risk. There is no fixed, guaranteed relationship, and therefore they cannot be perfect substitutes for money, i.e., money itself.

It is also a radical step to include cash surrender values of life insurance policies as part of the money supply. And yet, they too are balances that may be redeemed at any time, by promise of the life insurance company, that the policyholder cancels his policy. Like savings deposits, savings and loan shares, and savings bonds, they are considered by individuals as cash, and are valued as assets firmly at their cash value.⁴⁶ If savings deposits are accepted as part of the money supply, even though not directly used as media, then there is no cogent criterion for keeping out savings bonds and cash surrender values.⁴⁷ Yet those many economists who include savings deposits in the money supply have not yet pushed their logic to its final conclusion.⁴⁸

Much has been made of the legal permission to require notice for redeeming savings deposits and the other assets mentioned above. Yet everyone recognizes the economic fact that this provision is merely a dead letter; if notices were ever enforced, the bank would soon fail, as the enforcement would be considered by all as a sign of impending insolvency.⁴⁹ And the permitted notice requirements for the other assets are much shorter than the legal thirty days for savings deposits: life insurance surrender values and Series E savings bonds being practically immediate.

Neither is it permissible to distinguish between demand deposits and the other assets on the grounds that demand deposits do not pay interest. In the first place, commercial banks did pay interest on demand deposits until 1933, when the practice was outlawed.⁵⁰ And secondly, life insurance policies also do not obtain interest on the cash surrender values for the policyholder.

Thus, economists have a choice: they may either adopt the coherent but inexpedient definition of “money” as the narrow supply of legal tender; or, if they broaden their definition to include perfect money-substitutes, they should proceed onward to include all such assets, as outlined above.

Even if we adopt the latter course, however, we must still recognize the particularly strategic role of demand deposits as the direct medium. Here the analyses by McKinley and Shelby of the inverted money pyramids come into play. While savings banks, life insurance companies, etc. add to the money supply, they also keep the great bulk of their reserves in demand deposits rather than in cash, precisely because demand deposits are in such preponderant use as a direct medium. When we add up the total money supply outstanding in the hands of the public, then, we must not only deduct the cash reserves of the commercial banks, we must also deduct the demand-deposit reserves of these other money creators.⁵¹ And because these institutions keep their reserves in demand deposits, the Federal Reserve System, as the above authors have pointed out, exerts much greater control over them than purely legal considerations would lead us to believe.