1. Skip to navigation
  2. Skip to content
  3. Skip to sidebar

The Ludwig von Mises Institute

Advancing Austrian Economics, Liberty, and Peace

Advancing the scholarship of liberty in the tradition of the Austrian School

Search Mises.org

Part II
The Inflationary Boom: 1921-1929


The Inflationary Factors

Most writers on the 1929 depression make the same grave mistake that plagues economic studies in general—the use of historical statistics to "test" the validity of economic theory. We have tried to indicate that this is a radically defective methodology for economic science, and that theory can only be confirmed or refuted on prior grounds. Empirical fact enters into the theory, but only at the level of basic axioms and without relation to the common historical-statistical "facts" used by present-day economists. The reader will have to go elsewhere—notably to the works of Mises, Hayek, and Robbins—for an elaboration and defense of this epistemology. Suffice it to say here that statistics can prove nothing because they reflect the operation of numerous causal forces. To "refute" the Austrian theory of the inception of the boom because interest rates might not have been lowered in a certain instance, for example, is beside the mark. It simply means that other forces—perhaps an increase in risk, perhaps expectation of rising prices—were strong enough to raise interest rates. But the Austrian analysis, of the business cycle continues to operate regardless of the effects of other forces. For the important thing is that interest rates are lower than they would have been without the credit expansion. From theoretical analysis we know that this is the effect of every credit expansion by the banks; but statistically we are helpless—we cannot use statistics to estimate what the interest rate would have been. Statistics can only record past events; they cannot describe possible but unrealized events.

Similarly, the designation of the 1920s as a period of inflationary boom may trouble those who think of inflation as a rise in prices. Prices generally remained stable and even fell slightly over the period. But we must realize that two great forces were at work on prices during the 1920s—the monetary inflation which propelled prices upward and the increase in productivity which lowered costs and prices. In a purely free-market society, increasing productivity will increase the supply of goods and lower costs and prices, spreading the fruits of a higher standard of living to all consumers. But this tendency was offset by the monetary inflation which served to stabilize prices. Such stabilization was and is a goal desired by many, but it (a) prevented the fruits of a higher standard of living from being diffused as widely as it would have been in a free market; and (b) generated the boom and depression of the business cycle. For a hallmark of the inflationary boom is that prices are higher than they would have been in a free and unhampered market. Once again, statistics cannot discover the causal process at work.

If we were writing an economic history of the 1921-1933 period, our task would be to try to isolate and explain all the causal threads in the fabric of statistical and other historical events. We would analyze various prices, for example, to identify the effects of credit expansion on the one hand and of increased productivity on the other. And we would try to trace the processes of the business cycle, along with all the other changing economic forces (such as shifts in the demand for agricultural products, for new industries, etc.) that impinged on productive activity. But our task in this book is much more modest: it is to pinpoint the specifically cyclical forces at work, to show how the cycle was generated and perpetuated during the boom, and how the adjustment process was hampered and the depression thereby aggravated. Since government and its controlled banking system are wholly responsible for the boom (and thereby for generating the subsequent depression) and since government is largely responsible for aggravating the depression, we must necessarily concentrate on these acts of government intervention in the economy. An unhampered market would not generate booms and depressions, and, if confronted by a depression brought about by prior intervention, it would speedily eliminate the depression and particularly eradicate unemployment. Our concern, therefore, is not so much with studying the market as with studying the actions of the culprit responsible for generating and intensifying the depression-government.

The Definition of the Money Supply

Money is the general medium of exchange. On this basis, economists have generally defined money as the supply of basic currency and demand deposits at the commercial banks. These have been the means of payment: either gold or paper money (in the United States largely Federal Reserve Notes), or deposits subject to check at the commercial banks. Yet, this is really an inadequate definition. De jure, only gold during the 1920s and now only such government paper as Federal Reserve Notes have been standard or legal tender. Demand deposits only function as money because they are considered perfect money-substitutes, i.e., they readily take the place of money, at par. Since each holder believes that he can convert his demand deposit into legal tender at par, these deposits circulate as the unchallenged equivalent to cash, and are as good as money proper for making payments. Let confidence in a bank disappear, however, and a bank fail, and its demand deposit will no longer be considered equivalent to money. The distinguishing feature of a money-substitute, therefore, is that people believe it can be converted at par into money at any time on demand. But on this definition, demand deposits are by no means the only—although the most important—money-substitute. They are not the only constituents of the money supply in the broader sense.[1]

In recent years, more and more economists have begun to include time deposits in banks in their definition of the money supply. For a time deposit is also convertible into money at par on demand, and is therefore worthy of the status of money. Opponents argue (1) that a bank may legally require a thirty-day wait before redeeming the deposit in cash, and therefore the deposit is not strictly convertible on demand, and (2) that a time deposit is not a true means of payment, because it is not easily transferred: a check cannot be written on it, and the owner must present his passbook to make a withdrawal. Yet, these are unimportant considerations. For, in reality, the thirty-day notice is a dead letter; it is practically never imposed, and, if it were, there would undoubtedly be a prompt and devastating run on the bank.[2] Everyone acts as if his time deposits were redeemable on demand, and the banks pay out their deposits in the same way they redeem demand deposits. The necessity for personal withdrawal is merely a technicality; it may take a little longer to go down to the bank and withdraw the cash than to pay by check, but the essence of the process is the same. In both cases, a deposit at the bank is the source of monetary payment.[3] A further suggested distinction is that banks pay interest on time, but not on demand, deposits and that money must be non-interest-bearing. But this overlooks the fact that banks did pay interest on demand deposits during the period we are investigating, and continued to do so until the practice was outlawed in 1933.[4] Naturally, higher interest was paid on time accounts to induce depositors to shift to the account requiring less reserve.[5] This process has led some economists to distinguish between time deposits at commercial banks from those at mutual savings banks, since commercial banks are the ones that profit directly from the shift. Yet, mutual savings banks also profit when a demand depositor withdraws his account at a bank and shifts to the savings bank. There is therefore no real difference between the categories of time deposits; both are accepted as money-substitutes and, in both cases, outstanding deposits redeemable de facto on demand are many times the cash remaining in the vault, the rest representing loans and investments which have gone to swell the money supply.

To illustrate the way a savings bank swells the money supply, suppose that Jones transfers his money from a checking account at a commercial bank to a savings bank, writing a check for $1,000 to his savings account. As far as Jones is concerned, he simply has $1,000 in a savings bank instead of in a checking account at a commercial bank. But the savings bank now itself owns $1,000 in the checking account of a commercial bank and uses this money to lend to or invest in business enterprises. The result is that there are now $2,000 of effective money supply where there was only $1,000 before—$1,000 held as a savings deposit and another $1,000 loaned out to industry. Hence, in any inventory of the money supply, the total of time deposits, in savings as well as in commercial banks, should be added to the total of demand deposits.[6]

But if we concede the inclusion of time deposits in the money supply, even broader vistas are opened to view. For then all claims convertible into cash on demand constitute a part of the money supply, and swell the money supply whenever cash reserves are less than 100 percent. In that case, the shares of savings-and-loan associations (known in the 1920s as building-and-loan associations), the shares and savings deposits of credit unions, and the cash surrender liabilities of life insurance companies must also form part of the total supply of money.

Savings-and-loan associations are readily seen as contributing to the money supply; they differ from savings banks (apart from their concentration on mortgage loans) only in being financed by shares of stock rather than by deposits. But these "shares" are redeemable at par in cash on demand (any required notice being a dead letter) and therefore must be considered part of the money supply. Savings-and-loan associations grew at a great pace during the 1920s. Credit unions are also financed largely by redeemable shares; they were of negligible importance during the period of the inflationary boom, their assets totaling only $35 million in 1929. It might be noted, however, that they practically began operations in 1921, with the encouragement of Boston philanthropist Edward Filene.

Life insurance surrender liabilities are our most controversial suggestion. It cannot be doubted, however, that they can supposedly be redeemed at par on demand, and must therefore, according to our principles, be included in the total supply of money. The chief differences, for our purposes, between these liabilities and others listed above are that the policyholder is discouraged by all manner of propaganda from cashing in his claims, and that the life insurance company keeps almost none of its assets in cash—roughly between one and two percent. The cash surrender liabilities may be approximated statistically by the total policy reserves of life insurance companies, less policy loans outstanding, for policies on which money has been borrowed from the insurance company by the policyholder are not subject to immediate withdrawal.[7] Cash surrender values of life insurance companies grew rapidly during the 1920s.

It is true that, of these constituents of the money supply, demand deposits are the most easily transferred and therefore are the ones most readily used to make payments. But this is a question of form; just as gold bars were no less money than gold coins, yet were used for fewer transactions. People keep their more active accounts in demand deposits, and their less active balances in time, savings, etc. accounts; yet they may always shift quickly, and on demand, from one such account to another.

Inflation of the Money Supply, 1921-1929

It is generally acknowledged that the great boom of the 1920s began around July, 1921, after a year or more of sharp recession, and ended about July, 1929. Production and business activity began to decline in July, 1929, although the famous stock market crash came in October of that year. Table 1 depicts the total money supply of the country, beginning with $45.3 billion on June 30, 1921 and reckoning the total, along with its major constituents, roughly semiannually thereafter.[8] Over the entire period of the boom, we find that the money supply increased by $28.0 billion, a 61.8 percent increase over the eight-year period. This is an average annual increase of 7.7 percent, a very sizable degree of inflation. Total bank deposits increased by 51.1 percent, savings and loan shares by 224.3 percent, and net life insurance policy reserves by 113.8 percent. The major increases took place in 1922-1923, late 1924, late 1925, and late 1927. The abrupt leveling off occurred precisely when we would expect—in the first half of 1929, when bank deposits declined and the total money supply remained almost constant. To generate the business cycle, inflation must take place via loans to business, and the 1920s fit the specifications. No expansion took place in currency in circulation, which totaled $3.68 billion at the beginning, and $3.64 billion at the end, of the period. The entire monetary expansion took place in money-substitutes, which are products of credit expansion. Only a negligible amount of this expansion resulted from purchases of government securities: the vast bulk represented private loans and investments. (An "investment" in a corporate security is, economically, just as much a loan to business as the more short-term credits labeled "loans" in bank statements.) U.S. government securities held by banks rose from $4.33 billion to $5.50 billion over the period, while total government securities held by life insurance companies actually fell from $1.39 to $1.36 billion. The loans of savings-and-loan associations are almost all in private real estate, and not in government obligations. Thus, only $1 billion of the new money was not cycle-generating, and represented investments in government securities; almost all of this negligible increase occurred in the early years, 1921-1923.

See Table 1: Total Money Supply of the United States, 1921-1929 (in billions of dollars)

The other non-cycle-generating form of bank loan is consumer credit, but the increase in bank loans to consumers during the 1920s amounted to a few hundred million dollars at most; the bulk of consumer credit was extended by non-monetary institutions.[9]

As we have seen, inflation is not precisely the increase in total money supply; it is the increase in money supply not consisting in, i.e., not covered by, an increase in gold, the standard commodity money. In discussions of the 1920s, a great deal is said about the "gold inflation," implying that the monetary expansion was simply the natural result of an increased supply of gold in America. The increase in total gold in Federal and Treasury reserves, however, was only $1.16 billion from 1921-1929. This covers only a negligible portion of the total monetary expansion—the inflation of dollars.

Specifically, Table 2 compares total dollar claims issued by the U.S. government, its controlled banking system, and the other monetary institutions (the total supply of money) with the total holdings of gold reserve in the central bank (the total supply of the gold which could be used to sustain the pledges to redeem dollars on demand). The absolute difference between total dollars and total value of gold on reserve equals the amount of "counterfeit" warehouse receipts to gold that were issued and the degree to which the banking system was effectively, though not de jure, bankrupt. These amounts are compared for the beginning and end of the boom period.

The total of uncovered, or "counterfeited," dollars increased from $42.1 to $68.8 billion in the eight-year period, an increase of 63.4 percent contrasting to an increase of 15 percent in the gold reserve. Thus, we see that this corrected measure of inflation yields an even higher estimate than before we considered the gold inflow. The gold inflow cannot, therefore, excuse any part of the inflation.

Generating the Inflation, I: Reserve Requirements

What factors were responsible for the 63 percent inflation of the money supply during the 1920s? With currency in circulation not increasing at all, the entire expansion occurred in bank deposits and other monetary credit. The most important element in the money supply is the commercial bank credit base. For while savings banks, saving and loan associations, and life insurance companies can swell the money supply, they can only do so upon the foundation provided by the deposits of the commercial banking system. The liabilities of the other financial institutions are redeemable in commercial bank deposits as well as in currency, and all these institutions keep their reserves in the commercial banks, which therefore serve as a credit base for the other money-creators.[10] Proper federal policy, then, would be to tighten monetary restrictions on commercial banks in order to offset credit expansion in the other areas; failing, that is, the more radical reform of subjecting all of these institutions to the 100 percent cash reserve requirement.[11]

What factors, then, were responsible for the expansion of commercial bank credit? Since banks were and are required to keep a minimum percentage of reserves to their deposits, there are three possible factors—(a) a lowering in reserve requirements, (b) an increase in total reserves, and (c) a using up of reserves that were previously over the minimum legal requirement.

On the problem of excess reserves, there are unfortunately no statistics available for before 1929. However, it is generally known that excess reserves were almost nonexistent before the Great Depression, as banks tried to keep fully loaned up to their legal requirements. The 1929 data bear out this judgment.[12] We can safely dismiss any possibility that resources for the inflation came from using up previously excessive reserves.

We can therefore turn to the other two factors. Any lowering of reserve requirements would clearly create excess reserves, and thereby invite multiple bank credit inflation. During the 1920s, however, member bank reserve requirements were fixed by statute as follows: 13 percent (reserves to demand deposits) at Central Reserve City Banks (those in New York City and Chicago); 10 percent at Reserve City banks; and 7 percent at Country banks. Time deposits at member banks only required a reserve of 3 percent, regardless of the category of bank. These ratios did not change at all. However, reserve requirements need not only change in the minimum ratios; any shifts in deposits from one category to another are important. Thus, if there were any great shift in deposits from New York to country banks, the lower reserve requirements in rural areas would permit a considerable net overall inflation. In short, a shift in money from one type of bank to another or from demand to time deposits or vice versa changes the effective aggregate reserve requirements in the economy. We must therefore investigate possible changes in effective reserve requirements during the 1920s.

Within the class of member bank demand deposits, the important categories, for legal reasons, are geographical. A shift from country to New York and Chicago banks raises effective reserve requirements and limits monetary expansion; the opposite shift lowers requirements and promotes inflation. Table 3 presents the total member bank demand deposits in the various areas in June, 1921, and in June, 1929, and the percentage which each area bore to total demand deposits at each date.

We see that the percentage of demand deposits at the country banks declined during the twenties, from 34.2 to 31.4, while the percentage at urban banks increased, in both categories. Thus, the shift in effective reserve requirements was anti-inflationary, since the urban banks had higher legal requirements than the country banks. Clearly, no inflationary impetus came from geographical shifts in demand deposits.

What of the relation between member and non-member bank deposits? In June, 1921, member banks had 72.6 percent of total demand deposits; eight years later they had 72.5 percent of the total. Thus, the relative importance of member and non-member banks remained stable over the period, and both types expanded in about the same proportion.[13]

Table 3

*Banking and Monetary Statistics (Washington, D.C.: Federal Reserve Board, 1943), pp. 73, 81, 87 93, 99. These deposits are the official "U.S. Government" plus "other demand" deposits. They are roughly equal to "net demand deposits." "Demand deposits adjusted" are a better indication of the money supply and are the figures we generally use, but they are not available for geographic categories.

The relation between demand and time deposits offers a more fruitful field for investigation. Table 4 compares total demand and time deposits:

Thus, we see that the 1920s saw a significant shift in the relative importance of demand and time deposits: demand deposits were 51.3 percent of total deposits in 1921, but had declined to 44.5 percent by 1929. The relative expansion of time deposits signified an important lowering of effective reserve requirements for American banks: for demand deposits required roughly 10 percent reserve backing, while time deposits needed only 3 percent reserve. The relative shift from demand to time deposits, therefore, was an important factor in permitting the great monetary inflation of the 1920s. While demand deposits increased 30.8 percent from 1921 to 1929, time deposits increased by no less than 72.3 percent!

Time deposits, during this period, consisted of deposits at commercial banks and at mutual savings banks. Mutual savings banks keep only time deposits, while commercial banks, of course, also provide the nation's supply of demand deposits. If we wish to ask to what extent this shift from demand to time deposits was deliberate, we may gauge the answer by considering the degree of expansion of time deposits at commercial banks. For it is the commercial banks who gain directly by inducing their customers to shift from demand to time accounts, thereby reducing the amount of required reserves and freeing their reserves for further multiple credit expansion. In the first place, time deposits at commercial banks were about twice the amount held at mutual savings banks. And further, commercial banks expanded their time deposits by 79.8 percent during this period, while savings banks expanded theirs by only 61.8 percent. Clearly, commercial banks were the leaders in the shift to time deposits.

This growth in time deposits was not accidental. Before the establishment of the Federal Reserve System, national banks were not legally permitted to pay interest on time deposits, and so this category was confined to the less important state banks and savings banks. The Federal Reserve Act permitted the national banks to pay interest on time deposits. Moreover, before establishment of the Federal Reserve System, banks had been required to keep the same minimum reserve against time as against demand deposits. While the Federal Reserve Act cut the required reserve ratio roughly in half, it reduced required reserves against time deposits to 5 percent and, in 1917, to 3 percent. This was surely an open invitation to the banks to do their best to shift deposits from the demand to the time category.

During the 1920s, time deposits increased most in precisely those areas where they were most active and least likely to be misconstrued as idle "savings." Table 5 presents the record of the various categories of time deposits. The least active time accounts are in savings banks, the most active in the large city commercial banks. Bearing this in mind, below are the increases over the period in the various categories:

Thus, we see that, unerringly, the most active categories of time deposits were precisely the ones that increased the most in the 1920s, and that this correlation holds for each category. The most active—the Central Reserve City accounts—increased by 450 percent.[14]

Generating the Inflation, II: Total Reserves

Two influences may generate bank inflation—a change in effective reserve requirements and a change in total bank reserves at the Federal Reserve Bank. The relative strength of these two factors in the 1920s may be gauged by Table 6.

Clearly, the first four years of this period was a time of greater monetary expansion than the second four. The member bank contribution to the money supply increased by $6.9 billion, or 37.1 percent, in the first half of our period, but only by $3.9 billion or 15.3 percent in the second half. Evidently, the expansion in the first four years was financed exclusively out of total reserves, since the reserve ratio remained roughly stable at about 11.5 : 1. Total reserves expanded by 35.6 percent from 1921 to 1925, and member bank deposits rose by 37.1 percent. In the later four years, reserves expanded by only 8.7 percent, while deposits rose by 15.3 percent. This discrepancy was made up by an increase in the reserve ratio from 11.7 : 1 to 12.5 : 1, so that each dollar of reserve carried more dollars in deposits. We may judge how important shifts in reserve requirements were over the period by multiplying the final reserve figure, $2.36 billion, by 11.6, the original ratio of deposits to reserves. The result is $27.4 billion. Thus, of the $29.4 billion in member bank deposits in June, 1929, $27.4 billion may be accounted for by total reserves, while the remaining $2 billion may be explained by the shift in reserves. In short, a shift in reserves accounts for $2 billion out of the $10.8 billion increase, or 18.5 percent. The remaining 81.5 percent of the inflation was due to the increase in total reserves.

Thus, the prime factor in generating the inflation of the 1920s was the increase in total bank reserves: this generated the expansion of the member banks and of the non-member banks, which keep their reserves as deposits with the member banks. It was the 47.5 percent increase in total reserves (from $1.6 billion to $2.36 billion) that primarily accounted for the 62 percent increase in the total money supply (from $45.3 to $73.3 billion). A mere $760 million increase in reserves was so powerful because of the nature of our governmentally controlled banking system. It could roughly generate a $28 billion increase in the money supply.

What then caused the increase in total reserves? The answer to this question must be the chief object of our quest for factors responsible for the inflationary boom. We may list the well-known "factors of increase and decrease" of total reserves, but with special attention to whether or not they can be controlled or must be uncontrolled by the Federal Reserve or Treasury authorities. The uncontrolled forces emanate from the public at large, the controlled stem from the government.

There are ten factors of increase and decrease of bank reserves.

  1. Monetary Gold Stock. This is, actually, the only uncontrolled factor of increase—an increase in this factor increases total reserves to the same extent. When someone deposits gold in a commercial bank (as he could freely do in the 1920s), the bank deposits it at the Federal Reserve Bank and adds to its reserves there by that amount. While some gold inflows and outflows were domestic, the vast bulk were foreign transactions. A decrease in monetary gold stock causes an equivalent decrease in bank reserves. Its behavior is uncontrolled—decided by the public—although in the long run, Federal policies influence its movement.

  2. Federal Reserve Assets Purchased. This is the preeminent controlled factor of increase and is wholly under the control of the Federal Reserve authorities. Whenever the Federal Reserve purchases an asset, whatever that asset may be, it can purchase either from the banks or from the public. If it purchases the asset from a (member) bank, it buys the asset and, in exchange, grants the bank an increase in its reserve. Reserves have clearly increased to the same extent as Federal Reserve assets. If, on the other hand, the Federal Reserve buys the asset from a member of the public, it gives a check on itself to the individual seller. The individual takes the check and deposits it with his bank, thus giving his bank an increase in reserves equivalent to the increase in Reserve assets. (If the seller decides to take currency instead of deposits, then this factor is exactly offset by an increase in money in circulation outside the banks—a factor of decrease.)

    Gold is not included among these assets; it was listed in the first category (Monetary Gold Stock) and is generally deposited in, rather than purchased by, the Federal Reserve Banks. The major assets purchased are "Bills Bought" and "U.S. Government Securities." U.S. Government Securities are perhaps the most publicized field of "open-market operations"; Federal Reserve purchases add to bank reserves and sales diminish them. Bills Bought were acceptance paper which the Federal Reserve bought outright in a policy of subsidy that practically created this type of paper de novo in the United States. Some writers treat Bills Bought as an uncontrolled factor, because the Federal Reserve announced a rate at which it would buy all acceptances presented to it. No law, however, compelled it to adopt this policy of unlimited purchase; it therefore must be counted as a pure creation of Federal Reserve policy and under its control.

  3. Bills Discounted by the Federal Reserve. These bills are not purchased, but represent loans to the member banks. They are rediscounted bills, and advances to banks on their IOUs. Clearly a factor of increase, they are not as welcome to banks as are other ways of increasing reserves, because they must be repaid to the System; yet, while they remain outstanding, they provide reserves as effectively as any other type of asset. Bills Discounted, in fact, can be loaned precisely and rapidly to those banks that are in distress, and are therefore a powerful and effective means of shoring up banks in trouble. Writers generally classify Bills Discounted as uncontrolled, because the Federal Reserve always stands ready to lend to banks on their eligible assets as collateral, and will lend almost unlimited amounts at a given rate. It is true, of course, that the Federal Reserve fixes this rediscount rate, and at a lower rate when stimulating bank borrowing, but this is often held to be the only way that the System can control this factor. But the Federal Reserve Act does not compel, it only authorizes, the Federal Reserve to lend to member banks. If the authorities want to exercise an inflationary role as "lender of last resort" to banks in trouble, it chooses to do so by itself. If it wanted, it could simply refuse to lend to banks at any time. Any expansion of Bills Discounted, then, must be attributed to the will of the Federal Reserve authorities.

    On the other hand, member banks themselves have largely controlled the speed of repayment of Reserve loans. When the banks are more prosperous, they generally reduce their indebtedness to the Federal Reserve. The authorities could compel more rapid repayment, but they have decided to lend freely to banks and to influence banks by changing its rediscount charges.

    To separate controlled from uncontrolled factors as best we can, therefore, we are taking the rather drastic step of considering any expansion of Bills Discounted as controlled by the government, and any reduction as being uncontrolled, and determined by the banks. Of course, repayments will be partly governed by the amount of previous debt, but this seems to be the most reasonable division. We must take this step, therefore, even though it complicates the historical record. Thus, if Bills Discounted increase by $200 million over some three-year period, we may call this a controlled increase of $200 million, if we consider only this overall record. On the other hand, if we break down the record from year to year, it may be that Bills Discounted first increased by $500 million, then were reduced by $400 million, and then increased again by $100 million the final year. When we consider a year-to-year basis, then, controlled increase of reserves for the three years was $600 million and uncontrolled decrease was $400 million. The finer we break down the record, therefore, the greater the extent both of controlled increases by the government, and of uncontrolled declines prompted by the banks. Perhaps the best way of resolving this problem is to break down the record to the most significant periods. It would be far simpler to lump all Bills Discounted as controlled and let it go at that, but this would distort the historical record intolerably; thus, in the early 1920s it would give the Federal Reserve an undeserved accolade for reducing member bank debts when this reduction was largely accomplished by the banks themselves.

    We may therefore divide Bills Discounted into: New Bills Discounted (controlled factor of increase) and Bills Repaid (uncontrolled factor of decrease).

  4. Other Federal Reserve Credit. This is largely "float," or checks on banks remaining temporarily uncollected by the Federal Reserve. This is an interest-free form of lending to banks and is therefore a factor of increase wholly controlled by the Federal Reserve. Its importance was negligible in the 1920s.

  5. Money in Circulation Outside the Banks. This is the main factor of decrease—an increase in this item decreases total reserves to the same extent. This is the total currency in the hands of the public and is determined wholly by the relative place people wish to accord paper money as against bank deposits. It is therefore an uncontrolled factor, decided by the public.

  6. Treasury Currency Outstanding. Any increase in Treasury currency outstanding is deposited with the Federal Reserve in the Treasury's deposit account. As it is spent on government expenditures, the money tends to flow back into commercial bank reserves. Treasury currency is therefore a factor of increase, and is controlled by the Treasury (or by Federal statute). Its most important element is silver certificates backed 100 percent by silver bullion and silver dollars.

  7. Treasury Cash Holdings. Any increase in Treasury cash holdings represents a shift from bank reserves, while a decline in Treasury cash is spent in the economy and tends to increase reserves. It is therefore a factor of decrease and is controlled by the Treasury.

  8. Treasury Deposits at the Federal Reserve. This factor is very similar to Treasury cash holdings; an increase in deposits at the Reserve represents a shift from bank reserves, while a decrease means that more money is added to the economy and swells bank reserves. This is, therefore, a factor of decrease controlled by the Treasury.

  9. Non-member Bank Deposits at the Federal Reserve. This factor acts very similarly to Treasury deposits at the Federal Reserve. An increase in non-member bank deposits lowers member bank reserves, for they represent shifts from member banks to these other accounts. A decline will increase member bank reserves. These deposits are mainly made by non-member banks, and by foreign governments and banks. They are a factor of decrease, but uncontrolled by the government.

  10. Unexpended Capital Funds of the Federal Reserve. They are capital funds of the Federal Reserve not yet expended in assets (largely bank premises and expenses of operation). This capital is drawn from commercial banks, and, therefore, if unexpended, is a withdrawal of reserves. This is almost always a negligible item; it is clearly under the control of the Federal Reserve authorities.

Summing up, the following are the factors of change of member bank reserves:[15]

Factors of Increase

Monetary Gold Stock...................................................uncontrolled
Federal Reserve Assets Purchased....................................controlled
Bills Bought
U.S. Government Securities
New Bills Discounted.......................................................controlled
Other Federal Reserve Credit..........................................controlled
Treasury Currency Outstanding.......................................controlled
Factors of Decrease
Outside Money in Circulation......................................uncontrolled
Treasury Cash Holdings...................................................controlled
Treasury Deposits at the Federal Reserve........................controlled
Unexpended Capital Funds of the Federal Reserve........controlled
Non-member Bank Deposits at the Federal Reserve........................uncontrolled
Bills Repaid.......................................................uncontrolled

An overall survey of the entire 1921-1929 period does not give an accurate picture of the broad forces behind the movements in total reserves. For a while total reserves continued to increase. There were continual fluctuations within the various categories, with some increasing and other decreasing in any one period and different factors predominating at different times. Tables 7 and 8 depict the forces causing changes in total reserves during the 1920s. Table 7 breaks down 1921-1929 into 12 subperiods, shows the changes in each causal factor, and the consequent changes in member bank reserves, for each subperiod. Table 8 transforms the data of Table 7 into per-month figures, thus enabling comparison of the relative rates of change for the various periods.

Member bank reserves totaled $1604 million on June 30, 1921, and reached $2356 million eight years later. Over the 12 subperiods, uncontrolled reserves declined by $1.04 billion, while controlled reserves increased by $1.79 billion. By themselves, then, uncontrolled factors were deflationary; the inflation was clearly precipitated deliberately by the Federal Reserve. The plea that the 1920s was simply a "gold inflation" that the Federal Reserve did not counter actively is finally exploded. Gold was never the major problem, and in not one subperiod did it provide the crucial factor in increasing reserves.

In the 12 subperiods, uncontrolled factors declined seven times and increased five times. Controlled factors, on the other hand, rose in eight periods and declined in four. Of the controlled factors, Bills Bought played a vital role in changing reserves in nine periods, Government Securities in seven, Bills Discounted in five, and Treasury Currency in three (the first three). If we add up, regardless of arithmetic sign, the total impact of each controlled factor on reserves over the twelve periods, we find Government Securities in the lead (with $2.24 billion), Bills Bought slightly behind ($2.16 billion), and New Discounts behind that ($1.54 billion).

At the start of the eight-year period, Bills Discounted totaled $1.75 billion, Bills Bought were $40 million, U.S. Government Securities held were $259 million, Treasury Currency Outstanding totaled $1.75 billion, Monetary Gold Stock was $3 billion, and Money in Circulation was $4.62 billion.

Table 7 and Table 8 are organized as follows. Bills Discounted, Bills Bought, Government Securities held by the Federal Reserve, and Other Credit constitute Federal Reserve Credit. Changes in Federal Reserve Credit (except for net reductions in Bills Discounted), plus changes in Treasury Currency, Treasury Cash, Treasury Deposits at the Federal Reserve, and Unexpended Capital Funds of the Reserve constitute the controlled changes in member bank reserves. Changes in Monetary Gold Stock, Money in Circulation, and Other Deposits at the Federal Reserve constitute the uncontrolled changes, and the resultant effect constitutes the changes in bank reserves. The arithmetic signs of the actual changes of factors of decrease are reversed to accord with their effects on reserves; thus, a reduction of $165 million in Money in Circulation from 1921-1929 is listed in the table as a change of plus 165 on reserves.

Any division into historical periods is to a degree arbitrary. Yet the divisions of Tables 7 and 8 were chosen because the author believes they accord best with the most significant subperiods of the 1920s, subperiods which differ too much to be adequately reflected in any overall assessment. The following are the unique characteristics of each of these subperiods.

  1. June 1921-July 1922 (dates are all end-of-the-month). Bills Discounted, which had been falling since 1920, continued a precipitate decline, from $1751 million in June 1921 to a bottom of $397 million in August 1922. Total Reserve Credit also fell to reach a bottom in July 1922, as did Money in Circulation, which reached its bottom in July 1922. July was therefore chosen as the terminal month.
  2. July 1922-December 1922. Total Reserve Credit climbed upward sharply, hitting a peak in December, as did total reserves. Bills Discounted reached a peak in November.
  3. December 1922-October 1923. Bills Discounted continued to climb, reaching a peak in October. In the meanwhile, U.S. Government Securities fell sharply to reach a trough of 92 million in October, a trough for the whole period.
  4. October 1923-June 1924. Bills Bought fell abruptly, to reach a trough in July. Total Reserve Credit reached a trough in June.
  5. June 1924-November 1924. Bills Discounted, which had been falling since October 1923, continued to fall, reaching bottom in November 1924. U.S. Government Securities climbed to a peak in the same month. The Monetary Gold Stock also reached a peak in November 1924. Bills Bought climbed to a peak in December 1924, as did total Reserve Credit and total reserves.
  6. November 1924-November 1925. Bills Discounted climbed again to a peak in November. U.S. Securities fell to a trough in October, and total Reserve Credit reached a peak in December.
  7. November 1925-October 1926. U.S. Government Securities reached a trough in October, and Bills Discounted a peak in the same month. Clearly, both items milled around during this period.
  8. October 1926-July 1927. Bills Bought fell to a trough in July, and Bills Discounted reached a trough in August. Total Reserve Credit reached its trough in May.
  9. July 1927-December 1927. U.S. Government Securities climbed to a peak in December, as did Bills Discounted, total Reserve Credit, and total reserves.
  10. December 1927-July 1928. Bills Bought fell to a low in July, as did U.S. Government Securities, total Reserve Credit, and total reserves. Bills Discounted climbed to a peak in August.
  11. July 1928-December 1928. Bills Bought reached a peak in December, as did total reserves, while Bills Discounted and Reserve Credit reached a peak in November.
  12. December 1928-June 1929. Concludes the period under study.

Using these subperiods and their changes, we may now analyze precisely the course of the monetary inflation in the 1920s.

In Period I (June 1921-July 1922) a superficial glance would lead one to believe that the main inflationary factor was the heavy gold inflow, and that the Federal Reserve simply did not offset this influx sufficiently. A deeper analysis, however, shows that the banks paid off their loans at such a rapid rate that uncontrolled factors fell by $303 million. If the government had remained completely passive, therefore, member bank reserves would have declined by $303 million. Instead, the government actively pumped in $462 million of new reserves, yielding a net increase of $157 million. (Subtraction differences come from rounding.) The major channels of this increase were purchase of U.S. Government Securities ($278 million), Treasury currency ($115 million), and Bills Bought ($100 million).

Period II (July 1922-December 1922) saw a rapid acceleration of the inflation of reserves. Increasing at an average rate of $12 million per month in Period I, reserves now increased at a rate of $35 million per month. Once again, uncontrolled factors declined by $295 million, but they were more than offset by increases in controlled reserves pumped into the economy. These consisted of Bills Discounted ($212 million), Bills Bought ($132 million), and Treasury Currency ($93 million).

Periods III and IV (December 1922-June 1924) saw the inflation come roughly to a halt. Reserves actually fell slightly (by $4 million per month) in Period III (December 1922-October 1923), and rose only slightly (by $6 million per month) in Period IV. Simultaneously, bank deposits remained about level, member bank demand deposits staying at about $13.5 billion. Total deposits and total money supply, however, rose more in this period, with banks shifting to time deposits to permit increases. (Demand deposits rose by $450 million from June 1923 to June 1924, but time deposits rose by $1.5 billion). Total money supply rose by $3 billion. The economy responded to the slowdown of inflation by entering upon a mild minor recession, from May 1923 to July 1924.

The slight fall in reserves during Period III was brought about by selling U.S. Government Securities (-$344 million) and reducing the amount of bills held (-$67 million). This, indeed, was a positive decline, more than offsetting uncontrolled factors, which had increased by $132 million. The decline in reserves would have been even more effective, if the Federal Reserve had not increased its discounts ($266 million) and Treasury currency had not increased ($47 million).

Period IV (October 1923-June 1924), however, began to repeat the pattern of Period I and resume the march of inflation. Uncontrolled factors this time fell by $149 million, but they were more than offset by a controlled increase of $198 million, led by the heavy purchase of government securities ($339 million)—the heaviest average monthly buying spree yet seen in the 1920s ($42.4 million).

Period V was the most rapid reserve inflation to date, overreaching the previous peak of late 1922. Reserves increased by $39.8 million per month. Once again, the inflation was deliberate, uncontrolled factors declining by $262 million, but offset by a deliberate increase of $461 million. The critical factors of inflation were Bills Bought ($277 million) and U.S. Securities ($153 million).

The pace of inflation was greatly slowed in the next three periods, but continued nevertheless. From December 31, 1924 to June 30, 1927, reserves increased by $750 million; demand deposits adjusted, of all banks, rose by $1.1 billion. But time deposits rose by $4.3 billion during the same period, underscoring the banks' ability to induce customers to shift from demand to time deposits, while savings-and-loan shares and life-insurance reserves rose by another $4.3 billion. In 1926, there was a decided slowing down of the rate of inflation of the money supply, and this led to another mild economic recession during 1926 and 1927.

In Period VI (November 1924-November 1925), a tendency of uncontrolled reserves to decline was again more than offset by an increase in controlled reserves; these were Bills Discounted ($446 million) and Bills Bought ($45 million).

Period VII (November 1925-October 1926) was the first time after Period III that uncontrolled factors acted to increase reserves. But, in contrast, this time, the Federal Reserve failed to offset these factors sufficiently, although the degree of inflation was very slight (only $2.4 million per month).

In Period VIII (October 1926-July 1927), the degree of inflation was still small, but, ominously, the Federal Reserve stoked the fires of inflation rather than checked them; controlled factors increased, as did the uncontrolled. The culprits this time were the U.S. Government's Securities ($91 million) and Other Credit ($30 million).

Period IX (July 1927-December 1927), was another period of accelerated and heavy inflation, surpassing the previous peaks of latter 1922 and 1924. The per-monthly reserve increase in latter 1927 was $42.0 million. Once again, uncontrolled factors declined, but were more than offset by a very large increase in controlled reserves, emanating from Bills Bought ($220 million), U.S. Government Securities ($225 million), and Bills Discounted ($140 million).

Period X was the sharpest deflationary period (in reserves) in the 1920s. Uncontrolled factors rose, but were more than offset by a controlled decrease. Bills Discounted rose ($409 million), but the deflationary lead, was taken by U.S. Government Securities (-$402 million) and Bills Bought (-$230 million). The decline of over $200 million in reserves generated a decline of about $600 million in member bank demand deposits. Time deposits rose by over $1 billion, however, and life-insurance reserves by $550 million, so that the total money supply rose substantially, by $1.5 billion, from the end of 1927 to mid-1928.

With the boom now well advanced in years, and developing momentum, it was imperative for the Fed to accelerate its deflationary pressure, if a great depression was to be avoided. The deflation of reserves in the first half of 1928, as we have seen, was not even sufficient to offset the shift to time deposits and the other factors increasing the money supply. Yet, disastrously, the Fed resumed its inflationary course in latter 1928. In Period XI, a tendency of uncontrolled reserves to decrease, was offset by a positive and deliberate increase ($364 million of controlled reserves, against -$122 million of uncontrolled). The culprit in this program was Bills Bought, which increased by $327 million, while all the other reserve assets were only increasing slightly. Of all the periods of the 1920s, Period XI saw the sharpest average monthly rise in Bills Bought ($65.4 million).

In the final Period XII, the tide, at last, definitely and sharply turned. Uncontrolled factors increased by $390 million, but were offset by no less than a $423 million decrease in controlled reserves, consisting almost wholly of a reduction of $407 million in Bills Bought. Total reserves fell by $33 million. Member bank demand deposits, which also reached a peak in December, 1928, fell by about $180 million. Total demand deposits fell by $540 million.

So far, we have seen no reason why this deflation should have had any greater effect than the deflation of Period X. Indeed, total reserves fell by only $33 million as against $228 million in the former period. Member bank deposits fell by less ($180 million as against $450 million), and total demand deposits fell by about the same amount ($540 million against $470 million). The crucial difference, however, is this: in Period X, time deposits rose by $1.1 billion, insuring a rise in the nation's total currency and deposits of $600 million. But in Period XII, time deposits, far from rising, actually fell by $70 million. Total deposits, therefore, fell by $510 million, while the total money supply rose very slightly, impelled by continued growth in life insurance reserves. Time deposits no longer came to the rescue, as in 1923 and 1928, and total money supply rose only from $73 billion at the end of 1928, to $73.26 billion in mid-1929. For the first time since June 1921, the money supply stopped increasing, and remained virtually constant. The great boom of the 1920s was now over, and the Great Depression had begun. The country, however, did not really discover the change until the stock market finally crashed in October.

Treasury Currency

An increase in Treasury currency played a considerable role in the inflation in the early years from 1921-1923. It is unusual for Treasury currency to change considerably, as we see from its behavior over the rest of the 1921-1929 period. The surprising increase in 1921-1923 consisted almost exclusively of silver certificates, representing silver bullion held in the Treasury at 100 percent of its value. (Of the $225 million increase in Treasury currency during Periods I-III, $211 million was silver certificates.) In 1918, the Pittman Act had permitted the United States government to sell silver to Britain as a wartime measure, and the silver stock of the Treasury, as well as the silver certificates based 100 percent upon them, was reduced as a consequence. In May, 1920, however, in accordance with its obligation under the Act to buy silver bullion at the inflated price of one dollar per ounce until its stock had been replenished, the Treasury began to buy silver bullion, and this subsidy to domestic silver miners swelled bank reserves. This silver purchase policy effectively ended by mid-1923. The Treasury was forced to embark upon the silver purchase program by the terms of the Pittman Act of 1918, the responsibility of the Wilson administration. The Harding administration, however, could have repealed the Pittman Act if it had had the desire to do so. It must therefore bear its share of the blame for the silver purchase policy.[16]

Bills Discounted

We have seen the important role played by discounted bills in spurring the inflation. In 1923, 1925, and 1928, bills discounted came to the rescue of the banks at periods when the Fed was trying to exert anti-inflationary pressure by selling government securities, and, in 1923 and 1928 at least, reducing its holding of acceptances. In each instance, bills discounted was responsible for continuing the inflationary surge. The main trouble lay in the Federal Reserve's assumption of the role of "lender of last resort," more or less passively waiting to grant discounts to any banks that apply. But this was a policy adopted by the Fed, and it could have been changed at any time. The Fed allowed itself to affect discounts merely by setting and changing its rediscount rate.

The bulk of discounts consisted of rediscounted business paper (including commercial, agricultural, and industrial), and advances to banks on their promissory notes secured by U.S. government securities as collateral. When our period began, maximum legal maturity on discounts was 90 days, except for agricultural paper, which could be discounted for six months. In March, 1923, Congress extended the special privilege to agricultural paper to nine months, and the Fed was also granted authority to rediscount agricultural paper held by the newly established Federal Intermediate Credit Banks. More important, the FRB changed its original idea of making careful credit analyses of the original borrowers, and instead relied on the apparent solvency of the discounting banks, or else directly bailed out banks in distress.[17] This relaxation permitted a greater quantitative level of rediscounts.

If the Federal Reserve induced changes in discounts through the rediscount rate, it should certainly have always set it at a "penalty rate," i.e., high enough so that the banks would lose money by borrowing from it. If a bank earns 5 percent on its loan or investment, for example, and the Reserve sets its rediscount rate above that, say at 8 percent, then a bank will only borrow in the direst emergency when it desperately needs reserves. On the other hand, if the rediscount rate is set below the market, the bank can make a pleasant career out of borrowing, say, at 4 percent and relending the money at 5 percent. To discourage bank discounting, then, a permanent penalty rate above the market is essential. There was considerable opinion in the early 1920s that the FRB should maintain penalty rates in accord with British central banking tradition, but unfortunately the proponents only wanted rates above the lowest-yielding loans—prime commercial paper. Such a penalty rate would have been rather ineffectual, since the banks could still profit by discounting and relending to their riskier borrowers. A truly effective penalty rate would keep the rediscount rate above the rates of all bank loans.

Opinions clashed within the government in the early years on proposals for a mild penalty rate above prime commercial paper. The three main centers of monetary power were the Treasury, the Federal Reserve Board, and the New York Federal Reserve Bank, the latter two institutions clashing over power and policy throughout our period. At first, the Federal Reserve leaders favored penalty rates, and the Treasury was opposed: thus, the annual Federal Reserve Board report of 1920 promised establishment of the high rates.[18] By mid-1921, however, the Federal Reserve began to weaken, with Governor W.P.G. Harding, Chairman of the Federal Reserve Board, shifting his views largely for political reasons. Benjamin Strong, very powerful Governor of the Federal Reserve Bank of New York, also changed his mind at about the same time, and, as a result, penalty rates were doomed, and were no longer an issue from that point on.

Another problem of discount policy was whether the Federal Reserve should lend continuously to banks or only in emergencies.[19] While anti-inflationists must frown on either policy, certainly a policy of continuous lending is more inflationary, since it stokes the fires of monetary expansion continuously. The original theory of the Federal Reserve was to promote continuous credit, but for a while in the early 1920s, the Reserve shifted to favoring emergency credit only. Indeed, in an October, 1922 conference, FRB authorities approved the proposal of New York Federal Reserve Bank official, Pierre Jay, that the Federal Reserve should only supply seasonal and emergency credit and currency, and that even this should be restrained by the necessity of preventing credit inflation. By early 1924, however, the Federal Reserve abandoned this doctrine, and its Annual Report of 1923 supported the following disastrous policy:

The Federal Reserve banks are the . . . source to which the member banks turn when the demands of the business community have outrun their own unaided resources. The Federal reserve supplies the needed additions to credit in times of business expansion and takes up the slack in times of business recession.[20]

If the Federal Reserve is to extend credit during a boom and during a depression, it follows quite clearly that the Reserve's policy was frankly to promote continuous and permanent inflation.

Finally, in early 1926, Pierre Jay himself repudiated his own doctrine, and the "emergency" theory was now dead as a dodo.

Not only did the FRB, throughout the 1920s, keep rediscount rates below the market and lend continuously, it also kept delaying much needed raises in the rediscount rate. Thus, in 1923 and in 1925 the Fed sabotaged its own attempts to restrict credit by failing to raise the rediscount rate until too late, and it also failed to raise the rate sufficiently in 1928 and 1929.[21] One of the reasons for this failure was the Federal Reserve's consistent desire to supply "adequate" credit to business, and its fear of penalizing "legitimate business" through raising rates of interest. As soon as the Fed was established, in fact, Secretary of the Treasury William G. McAdoo trumpeted the policy which the Federal Reserve was to continue pursuing throughout the 1920s and during the Great Depression:

The primary purpose of the Federal Reserve Act was to alter and strengthen our banking system that the enlarged credit resources demanded by the needs of business and agricultural enterprises will come almost automatically into existence and at rates of interest low enough to stimulate, protect and prosper all kinds of legitimate business.[22]

Thus did America embark on its disastrous twentieth-century policy of inflation and subsequent depression—via a stimulation of legalized counterfeiting for special privilege conferred by government on favored business and farm enterprises.

As early as 1915 and 1916, various Board Governors had urged banks to discount from the Federal Reserve and extend credit, and Comptroller John Skelton Williams urged farmers to borrow and hold their crops for a higher price. This policy was continued in full force after the war. The inflation of the 1920s began, in fact, with an announcement by the Federal Reserve Board (FRB) in July, 1921, that it would extend further credits for harvesting and marketing in whatever amounts were legitimately required. And, beginning in 1921, Secretary of Treasury Andrew Mellon was privately urging the Fed that business be stimulated, and discount rates reduced; the records indicate that his advice was heeded to the full. Governor James, of the FRB, declared to his colleagues in 1926 that the "very purpose" of the Federal Reserve System "was to be of service to the agriculture, industry and commerce of the nation," and no one was apparently disposed to contradict him. Also in 1926, Dr. Oliver M.W. Sprague, economist and influential advisor to the Federal Reserve System, prophesied no immediate advances in the rediscount rate, because business had naturally been assuming since 1921 that plenty of Federal Reserve credit would always be available. Business, of course, could not be let down.[23] The Federal Reserve's very weak discount policy in 1928 and 1929 was caused by its fear that a higher interest rate would no longer "accommodate" business sufficiently.

An inflationary, low-discount-rate policy was a prominent and important feature of the Harding and Coolidge administrations. Even before taking office, President Harding had urged reduction of interest rates, and he repeatedly announced his intention of reducing discount rates after he became President. And President Coolidge, in a famous pre-election speech on October 22, 1924, declared that "It has been the policy of this administration to reduce discount rates," and promised to keep them low. Both Presidents appointed FRB members who favored this policy.[24] Eugene Meyer, chairman of the War Finance Corporation, warned the banks that by advertising that they do not discount with this farm loan agency, they were being "injurious to the public interest."[25] While such men as the head of the Merchants' Association of New York warned Coolidge about Federal Reserve credit to farmers, others pressed for more inflation: a Nebraska congressman proposed loans in new Treasury Notes at one-half percent to farmers, Senator Magnus Johnson urged a maximum rediscount rate of 2 percent, and the National Farmer-Labor Party called for the nationalization of all banking. Driven by their general desire to provide cheap and abundant credit to industry, as well as their policy (as we shall see below) of helping Britain avoid the consequences of its own monetary policies, the Federal Reserve sought constantly to avoid raising discount rates. In latter 1928 and 1929, with the need clearly evident, the FRB took refuge in the dangerous qualitative doctrine of "moral suasion." Moral suasion was an attempt to keep credit abundant to "legitimate" industry, while denying it to "illegitimate" stock market speculators. As we have seen, such attempts to segregate credit markets were inevitably self defeating, and were mischievous in placing different ethical tags on equally legitimate forms of business activity.

Moral suasion emerged in the famous February, 1929, letter of the FRB to the various Federal Reserve Banks, warning them that member banks were beyond their rights in making speculative loans, and advising restraint of Federal Reserve credit speculation, while maintaining credit to commerce and business. This step was taken in evasive response to persistent urging by the New York Federal Reserve Bank to raise the rediscount rate from 5 to 6 percent, a feeble enough step that was delayed until the latter part of 1929. Whereas, the New York Bank was the more inflationary organ in 1927 (as we shall see below), after that the New York Bank pursued a far more sensible policy: general credit restraint, e.g., raising the rediscount rate, while the Federal Reserve Board fell prey to qualitative credit fallacies at a peculiarly dangerous period—1929. The FRB went so far as to tell the New York Bank to lend freely and abundantly for commercial purposes.[26] The late Benjamin Strong had always held that it was impossible to earmark bank loans, and that the problem was quantitative and not qualitative. The New York Bank continued to stress this view, and refused to follow the FRB directive, repeating that it should not concern itself with bank loans, but rather with bank reserves and deposits.[27] The refusal of the New York Bank to follow the FRB directive of moral suasion finally drew a letter from the FRB on May 1, listing certain New York member banks that were borrowing continuously from the Federal Reserve, and were also carrying "too many" stock loans, and requesting that the New York Bank deal with them accordingly. On May 11, the New York Bank flatly refused, reiterating that banks have a right to make stock loans, and that there was no way to determine which loans were speculative. By June 1, the FRB succumbed, and dropped its policy of moral suasion. It did not raise the rediscount rate until August, however.[28]

Apart from the actions of the New York Bank, the policy of moral suasion failed, even on its own terms, for non-bank lenders used their bank-derived funds to replace bank lenders in the stock market. This inevitable result surprised and bewildered the qualitativists, and the stock market boom continued merrily onward.[29]

While stock market loans are no worse than any other form of loan, and moral suasion was a fallacious evasion of the need for quantitative restriction, any special governmental support for a certain type of loan is important in two ways: (1) government encouragement of one type of loan is apt to swell the overall quantity of bank loans; and (2) it will certainly overstimulate the particular loan and add to its readjustment difficulties in the depression phase. We must therefore examine the important instances of particular governmental stimulation to the stock market in the 1920s. While not as important as the increase in reserves and the money supply, this special aid served to spur the quantitative increase, and also created particular distortions in the stock market which caused greater troubles in the depression.

One important aid to stock market inflation was the FRS policy of keeping call loan rates (on bank loans to the stock market) particularly low. Before the establishment of the Federal Reserve System, the call rate frequently rose far above 100 percent, but since its inception, the call rate never rose above 30 percent, and very rarely above 10 percent.[30] The call rates were controlled at these low levels by the New York Federal Reserve Bank, in close cooperation with, and at the advice of, a Money Committee of the New York Stock Exchange. The New York Fed also loaned consistently to Wall Street banks for the purpose of regulating the call rate.

Another important means of encouraging the stock market boom was a rash of cheering public statements, designed to spur on the boom whenever it showed signs of flagging. President Coolidge and Secretary of Treasury Mellon in this way acted as the leading "capeadores of Wall Street."[31] Thus, when the emerging stock market boom began to flag, in January, 1927, Secretary Mellon drove it onward. The subsequent spurt in February leveled off in March, whereupon Mellon announced the Treasury's intention to refinance the 4? percent Liberty Bonds into 3? percent notes the next November. He predicted lower interest rates (accurately, due to the subsequent monetary inflation) and urged low rates upon the market. The announcement drove stock prices up again during March. The boom again began to weaken in the latter part of March, whereupon Mellon once more promised continued low rediscount rates and pictured a primrose path of easy money. He said, "There is an abundant supply of easy money which should take care of any contingencies that might arise." Stocks continued upward again, but slumped slightly during June. This time President Coolidge came to the rescue, urging optimism upon one and all. Again the market rallied strongly, only to react badly in August when Coolidge announced he did not choose to run again. After a further rally and subsequent recession in October, Coolidge once more stepped into the breach with a highly optimistic statement. Further optimistic statements by Mellon and Coolidge trumpeting the "new era" of permanent prosperity repeatedly injected tonics into the market. The New York Times declared on November 16 that Washington was the source of most bullish news and noted the growing "impression that Washington may be depended upon to furnish a fresh impetus for the stock market."

Bills Bought-Acceptances

Table 7 and Table 8 show the enormous importance of Bills Bought in the 1920s. While purchase of U.S. securities has received more publicity, Bills Bought was at least as important and indeed more important than discounts. Bills Bought led the inflationary parade of Reserve credit in 1921 and 1922, was considerably more important than securities in the 1924 inflationary spurt, and equally important in the 1927 spurt. Furthermore, Bills Bought alone continued the inflationary stimulus in the fatal last half of 1928.

These Bills Bought were all acceptances (and almost all bankers' acceptances), and the Federal Reserve policy on acceptances was undoubtedly the most curious, and the most indefensible, of the whole catalog of Federal Reserve policies. As in the case of securities, acceptances were purchased on the open market, and thus provided reserves to banks outright with no obligation to repay (as in discounting). Yet while the FRS preserved its freedom of action in buying or selling U.S. securities, it tied its own hands on acceptances. It insisted on setting a very low rate on acceptances, thus subsidizing and indeed literally creating the whole acceptance market in this country, and then pledging itself to buy all the bills offered at that cheap rate.[32] The Federal Reserve thus arbitrarily created and subsidized an artificial acceptance market in the United States and bought whatever was offered to it at an artificially cheap rate. This was an inexcusable policy on two counts—its highly inflationary consequences, and its grant of special privilege to a small group at the expense of the general public.

In contrast to Europe, where acceptances had long been a widely used form of paper, the very narrow market for them in this country, and its subsidization by the FRS, led to the Reserve's becoming the predominant buyer of acceptances.[33] It was a completely Federal Reserve-made market, and used only in international trade, or in purely foreign transactions. In 1928 and 1929, banks avoided borrowing from the Fed by making acceptance loans instead of straight loans, thus taking advantage of the FRS market and cheap acceptance rates. When the Federal Reserve bought the acceptance, the bank now acquired a reserve less expensively than by discounting, and without having to repay. Hence the inflationary role of acceptances in 1929 and its sabotaging of other Federal Reserve attempts to restrain credit.

In addition to acceptances held by the FRS on its own account, it also bought a large amount of acceptances as agent for foreign Central Banks. Moreover, the Reserve's buying rate on acceptances for foreign account was lower than for its own, thus subsidizing these foreign governmental purchases all the more. These holdings were not included in "Bills Bought," but they were endorsed by the FRS, and, in times of crisis, such endorsement could become a liability of the Federal Reserve; it did in 1931. The Reserve's acceptances were purchased from member banks, non-member banks, and private acceptances houses—with the bills for foreign account bought entirely from the private dealers.[34]

The first big investment in acceptances came in 1922, coinciding with the FRB's allowing the New York Reserve Bank to control acceptance policy. Federal Reserve holdings rose from $75 million in January to $272 million in December of that year. Despite the fact that the Federal Reserve kept its buying rate on acceptances below its rediscount rate, Paul Warburg, America's leading acceptance banker and one of the founders of the Federal Reserve System, demanded still lower buying rates on acceptances.[35] Undersecretary of the Treasury Gilbert, on the other hand, was opposed to the specially privileged acceptance rates, but the Federal Reserve continued its policy of subsidy, directed largely by the New York Bank.[36] It was, indeed, only in the first half of 1929 that the Federal Reserve partially abandoned its subsidizing, and at least pushed its buying rate on acceptances above the rediscount rate, thereby causing a sharp reduction in its acceptance holdings. In fact, the decline in acceptances was almost the sole factor in the decline of reserves in 1929 that brought the great inflation of the 1920s to its end.

Why did the Federal Reserve newly create and outrageously subsidize the acceptance market in this country? The only really plausible reason seems to center around the role played by Paul M. Warburg, former German investment banker who came to America to become a partner of Kuhn, Loeb and Company, and be one of the founders of the Federal Reserve System. Warburg worked for years to bring the rather dubious blessings of central banking to the hitherto backward United States. After the war and during the 1920s, Warburg continued to be chairman of the highly influential Federal Advisory Council, a statutory group of bankers advising the Federal Reserve System. Warburg, it appears, was a principal beneficiary of the Federal Reserve's pampering of the acceptance market. From its inception in 1920, Warburg was Chairman of the Board of the International Acceptance Bank of New York, the world's largest acceptance bank. He also became a director of the important Westinghouse Acceptance Bank and of several other acceptance houses, and was the chief founder and Chairman of the Executive Committee of the American Acceptance Council, a trade association organized in 1919. Surely, Warburg's leading role in the Federal Reserve System was not unconnected with his reaping the lion's share of benefits from its acceptance policy. And certainly, there is hardly any other way adequately to explain the adoption of this curious program. Indeed, Warburg himself proclaimed the success of his influence in persuading the Federal Reserve to loosen eligibility rules for purchase of acceptances, and to establish subsidized rates at which the Federal Reserve bought all acceptances offered.[37] And finally, Warburg was a very close friend of Benjamin Strong, powerful ruler of the New York Bank which engaged in the subsidy policy.[38]

The Federal government progressively widened the scope of the acceptance market from the very inception of the Federal Reserve Act. Before then, national banks had been prohibited from purchasing acceptances. After the Act, banks were permitted to buy foreign trade acceptances up to a limit of 50 percent of a bank's capital and surplus. Subsequent amendments raised the limit to 100 percent of capital and surplus, and then 150 percent, and allowed other types of acceptances—"dollar exchange," and domestic acceptances. Furthermore, English acceptance practice had been strictly limited to documentary exchange, representing definite movements of goods. The Federal Reserve Board at first tried to limit acceptance to such exchanges, but in 1923 it succumbed to the pressure of the New York Reserve Bank and permitted "finance bills" without documents. Wider powers were also granted to the New York and other Reserve Banks in 1921 and 1922 to purchase purely foreign acceptances, and their permissible maturity was raised from three to six months. In 1923, as part of the agricultural credit program, the Fed was permitted to rediscount agricultural-based acceptances up to six months.[39] In 1927, bills were made eligible even if drawn after the goods had been moved.[40]

With the rules relaxed, purely foreign acceptances, representing goods stored in or shipped between foreign points, rose from nothing to the leading role in Federal Reserve acceptance holdings during the crucial 1928-1929 period. Foreign acceptance purchases played a large part, especially in the latter half of 1929, in frustrating all attempts to check the boom. Previous credit restrictions had been on the way to ending the inflationary boom in 1928. But in August, the Federal Reserve deliberately reversed its tight money policy on the acceptance market, and the Board authorized the Federal Reserve Banks to buy heavily in order to accommodate credit needs.[41] The reasons for this unfortunate reversal were largely general: the political pressure for easier credit in an election year, and the fear of repercussions on Europe of high interest rates in the United States, played the leading roles. But there was also a more specific cause connected with the foreign acceptance market.

In contrast to older types of acceptance, the purely foreign acceptances were bills representing stored goods awaiting sale, rather than goods in transit between specific buyers and sellers.[42] The bulk was used to finance the storage of unsold goods in Central Europe, particularly Germany.[43] How did this increase in the holding of German acceptances come about? As the result of a spectacular American boom in foreign loans, financed by new issues of foreign bonds. This boom flourished from 1924 on, reaching a peak in mid-1928. It was the direct reflection of American credit expansion, and particularly of the low interest rates generated by that expansion. As we shall see further below, this result was deliberately fostered by the Federal Reserve authorities. Germany was one of the leading borrowers on the American market during the boom. Germany was undoubtedly short of capital, bereft as she was by the war and then by her ruinous inflation, culminating in late 1923. However, the German bonds floated in the United States did not, as most people thought, rebuild German capital. For these loans were largely extended to German local and state governments, and not to private German business. The loans made capital even scarcer in Germany, for the local governments were now able to compete even more strongly with private business for factors of production.[44] To their great credit, many German authorities, and especially Dr. Hjalmar Schacht, head of the Reichsbank, understood the unsoundness of these loans, and they together with the American Reparations Agent, Mr. S. Parker Gilbert, urged the New York banking community to stop lending to German local governments.[45] But American investment bankers, lured by the large commissions on foreign government loans, sent hundreds of agents abroad to urge prospective borrowers to float loans on the American market. They centered their attention on Germany.[46]

The tide of foreign lending turned sharply after mid-1928. Rising interest rates in the United States, combined with the steep stock exchange boom, diverted funds from foreign bonds to domestic stocks. German economic difficulties aggravated the slump in foreign lending in late 1928 and 1929. In consequence, German banks, finding their clients unable to float new bonds in the United States, obtained loans in the form of acceptance credits from the New York Reserve Bank, to cover the cost of carrying unsold stocks of cotton, copper, flour, and other commodities in German warehouses.[47] Those American banks that served as agents of foreign banks sold great quantities of foreign (largely German) acceptances to other American banks and to the FRS.[48] This explains the rise in Reserve holdings of German acceptances.

Other acceptances flourishing in 1928 and 1929 represented domestic cotton and wheat awaiting export, and exchange bills providing dollars to South America. In early 1929, there was also a rash of acceptances based on the import of sugar from Cuba, in anticipation of a heavier American tariff on sugar.[49]

Not only did the Federal Reserve—in effect the New York Bank—subsidize the acceptance market, it also confined its subsidizing to a few large acceptance houses. It refused to buy any acceptances directly from business, insisting on buying them from acceptance dealers as intermediaries—thus deliberately subsidizing the dealers. Further, it only bought acceptances from the few dealers with a capital of one million dollars and over. Another special privilege was the Federal Reserve's increasing purchase of acceptances under repurchase agreements. In this procedure, the New York Bank agreed to buy acceptances from a few large and recognized acceptance dealers who had the option to buy them back in 15 days at a currently fixed price. Repurchase agreements varied from one-tenth to almost two-thirds of acceptance holdings.[50] All this tends to confirm our hypothesis of the Warburg role.

In short, the Federal Reserve granted virtual call loans to the acceptance dealers, as well as unrestricted access at subsidized rates and accorded these privileges to dealers who were not, of course, members of the Federal Reserve System. In fact, as unincorporated private bankers, the dealers did not even make public reports. So curiously jealous was the New York Bank of the secrecy of its favorites that it arrogantly refused to give a Congressional investigating committee either a list of the acceptance dealers from whom it had bought bills, or a breakdown of foreign acceptances by countries. The officials of the New York Bank were not cited for contempt by the committee.[51]

U.S. Government Securities

Member bank reserves increased during the 1920s largely in three great surges—one in 1922, one in 1924, and the third in the latter half of 1927. In each of these surges, Federal Reserve purchases of government securities played a leading role. "Open-market" purchases and sales of government securities only emerged as a crucial factor in Federal Reserve monetary control during the 1920s. The process began when the Federal Reserve tripled its stock of government securities from November, 1921, to June, 1922 (its holdings totaling $193 million at the end of October, and $603 million at the end of the following May). It did so not to make money easier and inflate the money supply, these relationships being little understood at the time, but simply in order to add to Federal Reserve earnings. The inflationary result of these purchases came as an unexpected consequence.[52] It was a lesson that was appreciatively learned and used from then on.

If the Reserve authorities had been innocent of the consequences of their inflationary policy in 1922, they were not innocent of intent. For there is every evidence that the inflationary result was most welcome to the Federal Reserve. Inflation seemed justified as a means of promoting recovery from the 1920-1921 slump, to increase production and relieve unemployment. Governor Adolph Miller, of the Federal Reserve Board, who staunchly opposed the later inflationary policies, defended the 1922 inflation in Congressional hearings. Typical of Federal Reserve opinion at this time was the subsequent apologia of Professor Reed, who complacently wrote that bank credit "was being productively employed and that goods were being prepared for the consumer at least as rapidly as his money income was expanding."[53]

Open-market policy was then well launched, and played a major role in the 1924 and 1927 inflationary spurts and therefore in the overall inflation of the 1920s.

The individual Reserve Banks at first bought the securities on their own initiative, and this decentralized policy was resented by the Treasury. On the initiative of the Treasury, and seconded by Benjamin Strong, the Governors of the various Reserve Banks formed an Open-Market Committee to coordinate Reserve purchases and sales. The Committee was established in June, 1922. In April, 1923, however, this Governors' Committee was dissolved and a new Open-Market Investment Committee was appointed by the Federal Reserve Board. Originally, this was a coup by the Board to exert leadership over open-market policy in place of the growing power of Strong, Governor of the New York Bank. Strong was ill throughout 1923, and it was during that year that the Board managed to sell most of the FRB holdings of government securities. As soon as he returned to work in November, however, Strong, as chairman of the Open-Market Investment Committee, urged purchases of securities without hesitation should there be even a threat of business recession.

As a result of Strong's new accession to power, the Federal Reserve resumed within two months a heavy purchase of government securities, and the economy was well launched on its dangerous inflationary path. As Strong's admiring biographer puts it: "This time the Federal Reserve knew what it was doing, and its purchases were not earnings but for broad policy purposes," i.e., for inflation. Ironically, Benjamin Strong had now emerged as more powerful than ever, and in fact from that time until his retirement, the FRS's open-market policy was virtually controlled by Governor Strong.[54] One of Strong's first control devices was to establish a "Special System Investment Account," under which, as in the case of acceptances, Reserve purchases of governments were made largely by the New York Bank, which then distributed them pro rata to those other Reserve Banks that wanted the securities.

Another new and important feature of the 1920s was the maintenance of a large volume of floating, short-term government debt. Before the war, almost all of the U.S. debt had been funded into long-term bonds. During the war, the Treasury issued a myriad of short-term bills, only partially funded at a later date. From 1922 on, one half to one billion dollars of short-term Treasury debt remained outstanding in the banks, and had to be periodically refinanced. Member banks were encouraged to carry as much of these securities as possible: the Treasury kept deposits in the banks, and they could borrow from the Federal Reserve, using the certificates as collateral. Federal open-market purchases also helped make a market in government securities at low interest rates. As a result, banks held more government debt in 1928 than they had held during the war. Thus the Federal Reserve, by employing various means to bolster the market for Federal floating debt, added to the impetus for inflation.[55]

  1. See Lin Lin, "Are Time Deposits Money?" American Economic Review (March, 1937): 76-86. Lin points out that demand and time deposits are interchangeable at par and in cash, and are so regarded by the public. Also see Gordon W. McKinley, "The Federal Home Loan Bank System and the Control of Credit," Journal of Finance (September, 1957): 319-32, and idem, "Reply," Journal of Finance (December, 1958): 545.

  2. Governor George L. Harrison, head of the Federal Reserve Bank of New York, testified in 1931 that any bank suffering a run must pay both its demand and savings deposits on demand. Any request for a thirty-day notice would probably cause the state or the Comptroller of Currency to close the bank immediately. Harrison concluded: "in effect and in substance these [time] accounts are demanded deposits." Charles E. Mitchell, head of the National City Bank of New York, agreed that "no commercial bank could afford to invoke the right to delay payment on these deposits." And, in fact, the heavy bank runs of 1931-1933 took place in time deposits as well as demand deposits. Senate Banking and Currency Committee, Hearings on Operations of National and Federal Reserve Banking Systems, Part I (Washington, D.C., 1931), pp. 36, 321-22; and Lin Lin, "Are Time Deposits Money?"

  3. Time deposits, furthermore, are often used directly to make payments. Individuals may obtain cashier's checks from the bank, and use them directly as money. Even D.R. French, who tried to deny that time deposits are money, admitted that some firms used time deposits for "large special payments, such as taxes, after notification to the bank." D.R. French, "The Significance of Time Deposits in the Expansion of Bank Credit, 1922-1928," Journal of Political Economy (December, 1931): 763. Also see Senate Banking-Currency Committee, Hearings, pp. 321-22; Committee on Bank Reserves, "Member Bank Reserves" in Federal Reserve Board, 19th Annual Report, 1932 (Washington, D.C., 1933), pp. 27ff; Lin Lin, "Are Time Deposits Money?" and Business Week (November 16, 1957).

  4. See Lin Lin, "Professor Graham on Reserve Money and the One Hundred Percent Proposal," American Economic Review (March, 1937): 112-13.

  5. As Frank Graham pointed out, the attempt to maintain time deposits as both a fully liquid asset and an interest-bearing investment is trying to eat one's cake and have it too. This applies to demand deposits, savings-and-loan shares, and cash surrender values of life insurance companies as well. See Frank D. Graham, "One Hundred Percent Reserves: Comment," American Economic Review (June, 1941): 339.

  6. See McKinley, "The Federal Home Loan Bank System and the Control of Credit," pp. 323-24. On those economists who do and do not include time deposits as money, see Richard T. Selden, "Monetary Velocity in the United States," in Milton Friedman, ed., Studies in the Quantity Theory of Money (Chicago: University of Chicago Press, 1956), pp. 179-257.

  7. In his latest exposition of the subject, McKinley approaches recognition of the cash surrender value of life insurance policies as part of the money supply, in the broader sense. Gordon W. McKinley, "Effects of Federal Reserve Policy on Nonmonetary Financial Institutions," in Herbert V. Prochnow, ed., The Federal Reserve System (New York: Harper and Bros., 1960), pp. 217n., 222.

    In the present day, government savings bonds would have to be included in the money supply. On the other hand, pension funds are not part of the money supply, being simply saved and invested and not redeemable on demand, and neither are mutual funds—even the modern "open-end" variety of funds are redeemable not at par, but at market value of the stock.

  8. Data for savings-and-loan shares and life-insurance reserves are reliable only for the end-of-the-year: mid-year data are estimated by the author by interpolation. Strictly, the country's money supply is equal to the above data minus the amount of cash and demand deposits held by the savings and loan and life insurance companies. The latter figures are not available, but their absence does not unduly alter the results.

  9. On the reluctance of banks during this era to lend to consumers, see Clyde W. Phelps, The Role of the Sales Finance Companies in the American Economy (Baltimore, Maryland: Commercial Credit, 1952).

  10. As McKinley says:

    Just as the ultimate source of reserve for commercial banks consists of the deposit liabilities of the Federal Reserve Banks, so the ultimate source of the reserves of non-bank institutions consist of the deposit liabilities of the commercial banks. The money supply [is] . . . two inverted pyramids one on top of the other. The Federal Reserve stands at the base of the lower pyramid, and . . . by controlling the volume of their own deposit liabilities, the FRBs influence not only the deposit liabilities of the commercial banks but also the deposit liabilities of all those institutions which use the deposit liabilities of the commercial banks as cash reserves.

    "The Federal Home Loan Bank," p. 326. Also see Donald Shelby, "Some Implications of the Growth of Financial Intermediaries," Journal of Finance (December, 1958): 527-41.

  11. It might be asked, despairingly: if the supposedly "savings" institutions (savings banks, insurance companies, saving and loan associations, etc.) are to be subject to a 100 percent requirement, what savings would a libertarian society permit? The answer is: genuine savings, e.g., the issue of shares in an investing firm, or the sale of bonds or other debentures or term notes to savers, which would fall due at a certain date in the future. These genuinely saved funds would in turn be invested in business enterprise.

  12. Banking and Monetary Statistics, pp. 370-71. The excess listed for 1929 averages about forty million dollars, or about two percent of total reserve balances.

  13. Banking and Monetary Statistics, pp. 34 and 75. The deposits reckoned are "demand deposits adjusted" plus U.S. government deposits. A shift from member to non-member bank deposits would tend to reduce effective reserve requirements and increase excess reserves and the money supply, since non-member banks use deposits at member banks as the basis for their reserves. See Lauchlin Currie, The Supply and Control of Money in the United States (2nd ed., Cambridge, Mass.: Harvard University Press, 1935), p. 74.

  14. On time deposits in the 1920s, see Benjamin M. Anderson, Economics and the Public Welfare (New York: D. Van Nostrand, 1949), pp. 128-31; also C.A. Phillips, T.F. McManus, and R.W. Nelson, Banking and the Business Cycle (New York: Macmillan, 1937), pp. 98-101.

  15. The well-known category of "Federal Reserve Credit" consists of Federal Reserve Assets Purchased and Bills Discounted.

  16. For the Pittman Act, see Edwin W. Kemmerer, The ABC of the Federal Reserve System (9th ed., Princeton, N.J.: Princeton University Press, 1932), pp. 258-62.

  17. H. Parker Willis, "Conclusions," in H. Parker Willis, et al., "Report of an Inquiry into Contemporary Banking in the United States" (typewritten ms., New York, 1925), vol. 7, pp. 16-18.

  18. See Seymour E. Harris, Twenty Years of Federal Reserve Policy (Cambridge, Mass.: Harvard University Press, 1933), vol. 1, pp. 3-10, 39-48.

  19. Ibid., pp. 108ff.

  20. Federal Reserve, Annual Report, 1923, p. 10; cited in ibid., p. 109.

  21. See Phillips, et al., Banking and the Business Cycle, pp. 93-94.

  22. Harris, Twenty Years, p. 91.

  23. Oliver M.W. Sprague, "Immediate Advances in the Discount Rate Unlikely," The Annalist (1926): 493.

  24. See H. Parker Willis, "Politics and the Federal Reserve System," Banker's Magazine (January, 1925): 13-20; idem, "Will the Racing Stock Market Become A Juggernaut?" The Annalist (November 24, 1924): 541- 42; and The Annalist (November 10, 1924): 477.

  25. The War Finance Corporation had been dominant until 1921, when Congress expanded its authorized lending power and reorganized it to grant capital loans to farm cooperatives. In addition, the Federal Land Bank system, set up in 1916 to make mortgage loans to farm associations, resumed lending, and more Treasury funds for capital were authorized. And finally, the farm bloc pushed through the Agricultural Credits Act of 1923, which established twelve governmental Federal Intermediate Credit Banks to lend to farm associations. See Theodore Saloutos and John D. Hicks, Agricultural Discontent in the Middle West, 1900-1939 (Madison: University of Wisconsin Press, 1951), pp. 324-40.

  26. See Harris, Twenty Years, p. 209.

  27. Charles E. Mitchell, then head of the National City Bank of New York, has been pilloried for years for allegedly defying the FRB and frustrating the policy of moral suasion, by stepping in to lend to the stock market during the looming market crisis at the end of March. But it now appears that Mitchell and the other leading New York banks acted only upon approval of the Governor of the New York Federal Reserve Bank and of the entire Federal Reserve Board, which thus clearly did not even maintain the courage of its own convictions. See Anderson, Economics and the Public Welfare, p. 206.

  28. See Charles O. Hardy, Credit Policies of the Federal Reserve System (Washington, D.C.: Brookings Institution, 1932), pp. 122-38. Dr. Lawrence E. Clark, a follower of H. Parker Willis, charged that Mr. Gates McGarrah, Chairman of the New York Federal Reserve Bank at the time, opposed moral suasion because he himself was engaged in stock market speculation and in bank borrowing for that purpose. If this were the reason, however, McGarrah would hardly have been—as he was—the main force in urging an increase in the rediscount rate. Instead, he would have been against any check on the inflation. See Lawrence E. Clark, Central Banking Under the Federal Reserve System (New York: Macmillan, 1935), p. 267n.

  29. The moral suasion policy was searchingly criticized by former FRB Chairman W.P.G. Harding. The policy continued on, however, probably at the insistence of Secretary of the Treasury Mellon, who strongly opposed any increase in the rediscount rate. See Anderson, Economics and the Public Welfare, p. 210.

  30. See Clark, Central Banking, p. 382. The call rate rarely went above 8 percent in 1928, or above 10 percent in 1929. See Adolph C. Miller, "Responsibility for Federal Reserve Policies: 1927-1929," American Economic Review (September, 1935).

  31. Ralph W. Robey, "The Capeadores of Wall Street," Atlantic Monthly (September, 1928).

  32. Acceptances are sold by borrowers to acceptance dealers or "acceptance banks," who in turn sell the bills to ultimate investors—in this case, the Federal Reserve System.

  33. Thus, on June 30, 1927, over 26 percent of the nation's total of bankers' acceptances outstanding was held by the FRS for its own account, and another 20 percent was held for its foreign accounts (foreign central banks). Thus, 46 percent of all bankers' acceptances were held by the Federal Reserve, and the same proportion held true in June, 1929. See Hardy, Credit Policies, p. 258.

  34. See Senate Banking and Currency Committee, Hearings On Operation of National and Federal Reserve Banking Systems (Washington, D.C., 1931), Appendix, Part 6, p. 884.

  35. See Harris, Twenty Years, p. 324n.

  36. About half of the acceptances in the Federal Reserve System were held in the Federal Reserve Bank of New York; more important, almost all the purchases of acceptances were made by the New York Bank, and then distributed at definite proportions to the other Reserve Banks. See Clark, Central Banking, p. 168.

  37. See a presidential address by Warburg before the American Acceptance Council, January 19, 1923, in Paul M. Warburg, The Federal Reserve System (New York: Macmillan, 1930), vol. 2, p. 822. Of course, Warburg would have preferred an even larger subsidy. Even Warburg's perceptive warning on the developing inflation in March 1929, was marred by his simultaneous deploring of our "inability to develop a country-wide bill market." Commercial and Financial Chronicle (March 9, 1929): 1443-44; also see Harris, Twenty Years, p. 324.

  38. See Lester V. Chandler, Benjamin Strong, Central Banker (Washington, D.C.: Brookings Institution, 1958), p. 39 and passim. It was only on the insistence of Warburg and Henry Davison of J.P. Morgan and Company, that Strong had accepted this post.

  39. See H. Parker Willis, "The Banking Problem in the United States," in Willis, et al., "Report of an Inquiry into Contemporary Banking in the United States," pp. 1, 31-37.

  40. See A.S.J. Baster, "The International Acceptance Market," American Economic Review (June, 1937): 298.

  41. See Charles Cortez Abbott, The New York Bond Market, 1920-1930 (Cambridge, Mass.: Harvard University Press, 1937), pp. 124ff.

  42. See Hardy, Credit Policies, pp. 256-57. Also Hearings, Operation of Banking Systems, Appendix, Part C, pp. 852ff.

  43. Sterling bills were also purchased by the Fed to help Great Britain, e.g., $16 million in late 1929 and $10 million in the summer of 1927. See Hardy, Credit Policies, pp. 100ff.

  44. The boom in loans to Germany began with the 1924 "Dawes loan," part of the Dawes Plan reparations, with $110 million loaned to Germany by an investment banking syndicate headed by J.P. Morgan and Company.

  45. Schacht personally visited New York in late 1925 to press this course on the banks, and he, Gilbert, and German Treasury officials sent a cable to the New York banks in the same vein. The securities affiliate of the Chase National Bank did comply with these requests. See Anderson, Economics and the Public Welfare, pp. 150ff. See also Garet Garrett, A Bubble That Broke the World (Boston: Little, Brown, 1932), pp. 23-24, and Lionel Robbins, The Great Depression (New York: Macmillan, 1934), p. 64.

  46. "In late 1925, the agents of fourteen different American investment banking houses were in Germany soliciting loans from the German states and municipalities." Anderson, Economics and the Public Welfare, p. 152. Also see Robert Sammons, "Capital Movements," in Hal B. Lary and Associates, The United States in the World Economy (Washington, D.C.: U.S. Government Printing Office, 1943), pp. 95-100; and Garrett, A Bubble That Broke the World, pp. 20, 24.

  47. See Clark, Central Banking, p. 333. As early as 1924, the FRB had suggested that American acceptance credits finance the export of cotton to Germany.

  48. See H. Parker Willis, The Theory and Practice of Central Banking (New York: Harper and Bros., 1936), pp. 210-12, 223.

  49. Hearings, Operation of Banking Systems, pp. 852ff.

  50. Clark, Central Banking, pp. 242-48, 376-78; Hardy, Credit Policies, p. 248.

  51. Hearings, Operation of Banking Systems, Appendix, Part 6, pp. 847, 922-23.

  52. Yet not wholly unexpected, for we find Governor Strong writing in April, 1922 that one of his major reasons for open-market purchases was "to establish a level of interest rates . . . which would facilitate foreign borrowing in this country . . . and facilitate business improvement." Benjamin Strong to Under-Secretary of the Treasury S. Parker Gilbert, April 18, 1922. Chandler, Benjamin Strong, Central Banker, pp. 210-11.

  53. Harold L. Reed, Federal Reserve Policy, 1921-1930 (New York: McGraw-Hill, 1930), pp. 20, and 14-41. Governor Miller agreed "that though prices were moving upward, so was production and trade, and sooner or later production would overtake the rise of prices." Ibid., pp. 40-41.

  54. See Chandler, Benjamin Strong, Central Banker, pp. 222-33, esp. p. 233. Also see Hardy, Credit Policies, pp. 38-40; Anderson, Economics and the Public Welfare, pp. 82-85, 144-47.

  55. See H. Parker Willis, "What Caused the Panic of 1929?" North American Review (1930): 178; and Hardy, Credit Policies, p. 287. Tax exemption on income from government bonds also spurred the banks' purchases. See Esther Rogoff Taus, Central Banking Functions of the United States Treasury, 1789-1941 (New York: Columbia University Press, 1943), pp. 182ff.