The Fed’s monetary policy was consistently and unremittingly…
by Joseph T. Salerno via Mises Wire
In consecutive issues of The Freeman, Richard Timberlake has contributed an interesting trilogy of articles advancing a monetarist critique of the conduct of U.S. monetary policy during the 1920s and 1930s.1 In the first of these articles, Timberlake disputes the late Murray Rothbard’s “Austrian” account of the boom-bust cycle of the 1920s and 1930s. Timberlake contends that Rothbard proceeds on the basis of a “new and unacceptable meaning” for the term “inflation” and a contrived definition of the money supply to “invent” a Fed-orchestrated inflation of the 1920s that, in fact, never occurred. Moreover, Timberlake alleges, Rothbard’s account was marred by a “mismeasurement of the central bank’s monetary data” as well as by a misunderstanding of the nature and operation of the Fed-controlled pseudo-gold standard by which U.S. dollars were created during this period.
In the two subsequent articles, Timberlake also takes issue, respectively, with the U.S. Treasury’s policy of neutralizing gold inflows and the Fed’s policy of sharply raising reserve requirements in the mid-1930s, arguing that these complementary policies aborted an incipient economic recovery and brought on the recession of 1937–38. In what follows I will address the weighty charges brought against Rothbard and, in the process, offer an evaluation of the Federal Reserve System’s culpability for the economic events of these tragic years that diverges radically from Timberlake’s.
The Meaning of “Inflation”
Let me begin with Timberlake’s contention that Rothbard imputes a meaning to the word “inflation” that is both new and unacceptable. In fact Rothbard’s definition of inflation as “the increase in money supply not consisting in, i.e., not covered by, an increase in gold,” is an old and venerable one. It was the definition that was forged in the theoretical debate between the hard-money British Currency School and the inflationist British Banking School in the mid-nineteenth century. According to the proto-Austrian Currency School, which triumphed in the debate, the gold standard was not sufficient to prevent the booms and busts of the business cycle, which had continued to plague Great Britain despite its restoration of the gold standard in 1821.2
Briefly, according to the Currency School, if commercial banks were permitted to issue bank notes via lending or investment operations in excess of the gold deposited with them this would increase the money supply and precipitate an inflationary boom. The resulting increase in domestic money prices and incomes would eventually cause a balance-of-payments deficit financed by an outflow of gold. This external drain of their gold reserves and the impending threat of internal drains due to domestic bank runs would then induce the banks to sharply restrict their loans and investments, resulting in a severe contraction of their uncovered notes or “fiduciary media” and a decline in the domestic money supply accompanied by economy-wide depression.
To avoid the recurrence of this cycle, the Currency School recommended that all further issues of fiduciary media be rigorously suppressed and that, henceforth, the money supply change strictly in accordance with the inflows and outflows of gold through the nation’s balance of payments. The latter provided a natural, noncycle-generating mechanism for distributing the world’s money supply strictly in accordance with the international pattern of monetary demands.
Following the triumph of the Currency School doctrine and the implementation of its policy prescription by the Bank of England, its definition of inflation became accepted in the English-speaking world, especially in the United States, where there existed a much more radical and analytically insightful American branch of the School. The term “inflation” was now used strictly to denote an increase in the supply of money that consisted in the creation of currency and bank deposits unbacked by gold. Thus for example, the American financial writer Charles Holt Carroll wrote in 1868 that “The source of inflation, and of the commercial crisis, is in the nature of the system which pretends to lend money, but creates currency by discounting such bills when there is no such money in existence.”3 Even earlier, in 1858, Carroll had written, “Instead of using gold and silver for currency they are merely used as the basis of the greatest possible inflation by the banks,” and that “we should prevent any artificial increase of currency to prevent a future … catastrophe.”4 So it was the “artificial increase of currency” only — through the creation of unbacked bank notes and deposits — that constituted inflation.
The leading monetary theorist in the United States in the last quarter of the nineteenth century was Francis A. Walker. According to Walker, writing in 1888, “A permanent excess of the circulating money of a country, over that country’s distributive share of the money of the commercial world is called inflation.”5 While this version of the definition is applicable to inconvertible paper fiat currency, Walker also believed that inflation was an inherent feature of the issuance of convertible bank notes and deposits that lacked gold backing. In Walker’s words, “there resides in bank money, even under the most stringent provisions for convertibility, the capability of local and temporary inflation.”6
Unfortunately, however, because the writers of the British Currency School, unlike their American cousins, neglected to consider bank deposits as part of the money supply, their policies as adopted in Great Britain failed to prevent inflation and the business cycle. Consequently, and tragically, the School’s doctrines and policies fell into profound disrepute by the late nineteenth century, and its definition of inflation was replaced by that of the opposing Banking School, which saw inflation as a state in which the money supply exceeds the needs of trade.
Early American quantity theorists following the proto-monetarist Irving Fisher, in particular, seized upon and adapted this definition to their peculiar analytical perspective. Thus, Edwin Kemmerer wrote in 1920 that, “Although the term inflation in current discussion is used in a variety of meanings, there is one idea common to most uses of the word, namely, the idea of a supply of circulating media in excess of trade needs.”7 Kemmerer went on to define inflation as a state in which, “at a given price level, a country’s circulating media — money and deposit currency — increase relatively to trade needs.” From here it was a short step to the currently prevailing definition of inflation as an increase in the price level.8
So Rothbard’s theory is surely not new and to say that it is “unacceptable” is simply to express one’s agreement with the long-entrenched preference among orthodox quantity theorists, including contemporary monetarists, for the Banking School over the Currency School.
Timberlake also challenges Rothbard’s statistical definition of the money supply for including savings and loan share capital and life insurance net policy reserves, alleging that Rothbard contrived this definition in order to make the rate of monetary growth appear larger than it actually was during the 1920s. Timberlake argues that the two items in question are not money because “they cannot be spent on ordinary goods and services. To spend them, one needs to cash them in for other money — currency or bank drafts.”9 Let us take these items one at time.
In the case of savings and loan share capital, there are two responses to Timberlake. First, the “share accounts” offered by savings and loan associations are and always have been economically indistinguishable from the savings deposits offered by commercial banks, included in the older (pre-1980) definition of M2 that Timberlake apparently upholds as the appropriate definition of the money supply.10 In practice depositors could withdraw their savings deposits from commercial banks on demand, because the law that permitted the banks to insist on a waiting period was rarely if ever invoked. Similarly, while savings and loan associations were contractually obligated to “repurchase” their “shares” at par on request of the shareholder, they could legally delay such repurchase for shorter or longer periods depending on their individual bylaws. Nonetheless such delays rarely occurred and “for many years savings and loan associations have made the proud boast ‘every withdrawal paid upon demand’ or some similar statement.11
Moreover, while Timberlake is right that “shareholders” had to trade their share accounts in for currency or bank drafts (at par and on demand) before they could spend them on goods and services, this was equally true of savings depositors at commercial banks. Thus the public has always considered dollars held in savings and loan share accounts or savings accounts as readily spendable as dollars held in commercial bank savings deposits.
Second, Timberlake curiously does not object to Rothbard’s inclusion of the savings deposits of mutual savings banks in the money supply, although they also are not included in the M2 definition he favors.12 What makes Timberlake’s position even more puzzling is that mutual savings banks were practically identical in economic function to savings and loan associations and were also technically “mutually” owned by their depositors.13 So why, then, does Timberlake insist so vehemently on treating the liabilities of these two institutions differently?
A resolution of this mystery can perhaps be found in the work of Milton Friedman and Anna Schwartz, who excluded the share accounts of savings and loans (and of credit unions) from their definition of the money supply on the grounds that these institutions are technically not banks as defined “in accordance with the definition of banks agreed upon by federal bank supervisory agencies” since “holders of funds in these institutions are for the most part technically shareholders, not depositors.” Despite this legal technicality, however, even Friedman and Schwartz were forced to admit that those who place funds with these institutions “clearly … may regard such funds as close substitutes for bank deposits, as we define them.”14
Life Insurance Reserves
This brings us to the issue of the net policy reserves of life insurance companies. Rothbard claimed that the cash surrender values of life insurance companies, that is, the immediately cashable claims possessed by policyholders against life insurance companies, statistically approximated by the companies’ net policy reserves, represent a source of currently spendable dollars and should be included in the money supply. Once again the question is not whether insurance companies superficially resemble banks or can be technically classified as such according to some arbitrary regulatory definition. It is whether they essentially function like depository institutions, receiving funds from the public with which to make loans and investments, while contractually promising that such funds are available for withdrawal on demand by the policyholder. In Rothbard’s view, the policyholder is economically in precisely the same position as a bank depositor (and thrift institution shareholder) in holding an immediately cashable par-value claim to dollars.
Now admittedly, Rothbard’s inclusion of this item in the money supply is controversial, much more so than his inclusion of savings and loan share accounts. However, he was hardly alone in maintaining this position. A number of mainstream writers of money and banking textbooks in the 1960s and 1970s recognized that cashable life insurance reserves possessed some of the characteristics of money. For example, Walter W. Haines characterized insurance companies as “savings institutions” and noted that these savings “can be withdrawn at any time” simply by allowing the policy to lapse, a feature that marks them as a “near-money” on a par with savings accounts.15 M.L. Burstein maintained that the cash value of a life insurance policy offered “ready convertibility” into cash, was “almost as liquid as a mattressful of currency,” and satisfied the “precautionary motive” for holding liquid assets no less than savings and loan accounts and savings bonds.16 Albert Hart and Peter Kenen included the “net cash values of life insurance” in the broadest class of financial assets possessing the attribute of “moneyness,” while Thomas F. Cargill ranked them on a liquidity spectrum immediately below large certificates of deposit, which are included in the current M3 definition of the money supply.17
More important, however, even if we grant for the sake of argument that net life insurance reserves should be excluded from the money supply, we find that it makes very little difference to Rothbard’s characterization of the 1920s as an inflationary decade. With this item included, the increase in Rothbard’s M between mid-1921 and the end of 1928 totaled about 61 percent, yielding an annual rate of monetary inflation of 8.1 percent a year; with this item left out (but savings and loan share accounts included), the money supply increased by about 55 percent over the period or at an annual rate of 7.3 percent.18 Mirabile dictu, by using a definition of the money stock that arbitrarily excludes savings and loan share accounts while including mutual savings bank deposits on the basis of an inexplicable adherence to a legalistic regulatory definition of banks, it turns out that it is Timberlake (and Friedman and Schwartz) who have mismeasured money supply growth during the 1920s.
Timberlake also criticizes Rothbard for “ignorance of the flawed institutional framework within which the gold standard and the central bank generated money” and also of “mismeasurement of the central bank’s monetary data.”19 But this is surely a curious charge to level against Rothbard, steeped as he was in Currency School doctrine. In fact, Rothbard was quite cognizant that the U.S. monetary regime of the 1920s and 1930s was not a genuine gold standard in which the supply of money was determined exclusively by market forces, that is, by the balance of payments and the mining of gold, but a hybrid system in which the Fed possessed substantial power to manipulate the money supply by pyramiding paper bank reserves atop its stock of gold reserves. Indeed, Rothbard went much further than Timberlake in rigorously and completely separating those factors affecting the money supply that were subject to Fed control from those that the Fed had no control over.20
In analyzing the central bank monetary data, Timberlake starts with the monetary base or “Total Fed,” which is equal to currency in circulation plus member bank reserves. From this aggregate he properly subtracts the Fed’s legal-tender reserves, mainly the gold stock, whose size depends on balance-of-payments flows and is not under the immediate control of the Fed. What remains is the “net monetary obligations” of the Fed or “Net Fed,” which, according to Timberlake, “faithfully indicates the intent of Fed policy.”21 From 1921 to 1929, this aggregate declined by 8 percent per year, leading Timberlake to conclude that the intent of Fed policy was decidedly deflationary during this period. The motive for this deflationary policy bias was, Timberlake suggests, to aid Great Britain in re-establishing and maintaining gold convertibility for the pound sterling.
However, as important as it is, the gold stock is not the only factor that lay beyond the Fed’s control. For as Rothbard points out, currency in circulation, which improperly remains in Timberlake’s Net Fed aggregate, is not controlled by the Fed at all but by the banking public. Any time a depositor withdraws cash from a bank, currency in circulation increases and bank reserves decline, dollar for dollar. Under a fractional-reserve banking system, this loss of reserves causes a multiple contraction of bank deposits that far exceeds the original increase in currency in circulation that induced it and therefore results in a net deflation of the money supply. Conversely, a decline in the amount of currency held by the public causes an overall increase in bank reserves and an overall inflation of the money supply.
This is not all, however — Timberlake also ignores the fact that under the prevailing policy regime the banks themselves could autonomously reduce the amount of bank reserves and thus the quantity of money in existence by deliberately reducing their indebtedness to the Fed. During this period, it was the chosen policy of the Fed to lend liberally and continuously to all banks at an interest, or “discount,” rate below the market rate. While the Fed was legally authorized to make such loans to its member banks, it was not mandated to do so. Furthermore, it also retained complete power to set the “discount rate” it charged on these loans. Hence, if it had chosen to, the Fed could have restricted its lending to emergency situations and charged a penalty rate substantially above the market rate, so as to discourage all but the most seriously troubled banks from applying for loans. In short, it could have almost completely neutralized the inflationary impact of its discounting operations. This “emergency lending” policy had been urged by some prominent officials within the Fed establishment itself.22
The fact that the Fed chose instead to pursue a “continuous lending” policy meant that the increase in bank reserves that resulted from the origination of new Fed loans to member banks via the rediscounting of business bills or advances on collateralized bank promissory notes was under the exclusive control of the Fed. But it also meant that the reduction in bank reserves entailed by the net repayment of discounted bills was uncontrolled by the Fed, because it depended solely on the decisions of the banks. Given the Fed’s indiscriminate, below-market rate discount policy, the banks were always in a position to maintain or augment their debts to the Fed if they so desired simply by discounting additional bills with the Fed. Thus, as Rothbard concluded, when “Bills Repaid” exceeded “New Bills Discounted,” banks were deliberately and autonomously diminishing their level of indebtedness to the Fed and this must be counted as an uncontrolled deflationary influence on bank reserves.
Real Fed Intent
If we follow Rothbard, then, in identifying currency in circulation and the reduction of bank indebtedness to the Fed along with the gold stock as the main “uncontrolled” factors affecting bank reserves, we get a picture of the Fed’s intent during the 1920s and early 1930s that is poles apart from the one suggested by Timberlake. Indeed, we find that from the inception of the monetary inflation in mid-1921 to its termination at the end of 1928, “uncontrolled reserves” decreased by $1.430 billion while controlled reserves increased by $2.217 billion. Since member bank reserves totaled $1.604 billion at the beginning of this period, this means that controlled reserves shot up by 138 percent or 18.4 percent per year during this seven-and-one-half year period, while uncontrolled reserves fell by 89 percent or 11.9 percent per year. Thus Rothbard correctly concluded that the 1920s were an inflationary decade and that it was indeed the intention of the Federal Reserve System that it be so.23
The Fed’s inflationary intent is perfectly consistent, moreover, with its motive of helping Great Britain re-establish and maintain the pre-war parity between gold and the British pound. While Timberlake properly recognizes this motive underlying Fed policy, he is incorrect in suggesting that it necessitates a deflationary policy on the part of the Fed. In fact, the precise opposite is required. The British pound in the mid-1920s was overvalued vis-à-vis gold and the U.S. dollar, causing British products to appear relatively overpriced in world markets. As a result, Great Britain experienced imports chronically in excess of exports accompanied by persistent balance-of-payments deficits and outflows of gold reserves. Had the Fed deflated the U.S. money supply, thus lowering U.S. prices even more relative to British prices as Timberlake claims was its intention, it would have exacerbated, and not resolved, Great Britain’s gold drain. Clearly, then, the Fed’s desire to aid Britain in reversing its balance-of-payments deficits and rebuilding its gold stocks called for an inflationary policy intended to pump up U.S. prices, thereby rendering British products relatively cheap and enhancing the demand for them on world markets.24
This point about the motive for the Fed’s easy-money policy in the 1920s was not only advanced by Rothbard, but by other economists, including monetarists such as Kenneth Weiher. According to Weiher:
Great Britain was calling for help [in 1924] and Benjamin Strong [president of the New York Fed] heard the call. Expansionary monetary policy in the U. S. would drive prices up and interest rates down in this country, which would tend to send gold flowing toward Great Britain, where prices were lower and interest rates higher. These changes would help America’s ally build up its stock of gold. … [T]here can be no question that the Fed would not have moved when it did were it not for concern over the gold standard and the plight of Great Britain. … By 1927, the stagnant British economy needed help from the United States and the rest of Europe. … Just as had been the case in 1924, monetary policy was shifted to an expansionary program in an effort to aid Great Britain’s struggles to return to the gold standard.25
Rothbard’s reinterpretation of the monetary data also cuts against Timberlake’s claim that the Fed “monetarily starved the country into the worst economic crisis it has ever experienced.”26 On the contrary, the factors controlled by the Fed continued to exercise a highly inflationary impact on bank reserves and the money supply from late 1929 through 1932, as the Fed attempted desperately to ward off the depression precipitated by the termination of the bank credit inflation that it had orchestrated in the 1920s.
The deflation of the money supply, therefore, was caused wholly by factors beyond the control of the Fed. First, there was a loss of confidence in the Fed-dominated phony gold standard among the domestic public and foreign investors. As a result there occurred an increase in currency in circulation and a decline in the Fed’s gold stock, both of which caused bank reserves to decline. Second, U.S. banks prudently attempted to save themselves and their depositors by restricting their loans to overcapitalized and failing businesses and instead using these funds to pay down their indebtedness to the Fed, which gave further impetus to the “uncontrolled” reduction of bank reserves. Third, in the second quarter of 1932, the banks also began to increase their liquid reserves beyond the legal minimum. The accumulation of “excess reserves,” as they were called, constituted a separate uncontrolled factor that reinforced the deflationary influence of the uncontrolled decline in bank reserves on the money supply.
From the end of December 1929 to the end of December 1931, bank reserves fell from $2.36 billion to $1.96 billion causing RM (for Rothbard’s money supply) to drop from $73.52 billion to $68.25 billion or at an annual rate of 3.6 percent. But this monetary deflation was not caused by the Fed, which pumped up controlled reserves by $672 million or at an annual rate of 17 percent during the period, while uncontrolled reserves declined by $1,063 million or by 27 percent per year. During 1932, RM continued to decline, falling to $64.72 billion or by 5.2 percent. But bank reserves increased sharply during the year from $1.96 billion to $2.51 billion, as the Fed furiously inflated controlled reserves. In the last ten months of the year, controlled reserves rose by a staggering $1,165 million, or at an annual rate of 76 percent. Fortunately, this attempted massive inflation of the money supply was undone by the domestic public, foreign investors, and the banks as uncontrolled reserves dwindled by $495 million and banks began to accumulate substantial excess reserves.
The story was much the same in 1933 as a determined inflationary campaign conducted by the Fed in the early part of the year — controlled reserves rose by $785 million in February alone — was defeated by the public and the banks, and RM declined by over $3 billion, or by almost 5 percent.27
So once the data have been properly arranged and interpreted, it becomes clear that the Fed does not deserve praise for the bank credit deflation of 1930–1933. This honor goes to private dollar-holders, domes-tic and foreign, who attempted to reclaim their rightful property from a central bank-manipulated and inflationary financial system masquerading as a gold standard that had repeatedly betrayed their trust.
In two follow-up articles, Timberlake extends his attack on what he considers to be the “deflationary” monetary policies pursued by the Treasury and Fed in the mid-1930s. In particular, he criticizes the Treasury’s policy of “neutralizing,” or “sterilizing,” the effect of the inflow of gold on bank reserves from late 1936 to early 1938 and the Fed’s policy of increasing reserve requirements in 1936 and 1937. But neither of these policies caused a contraction of the money supply. They merely temporarily interrupted a massive monetary inflation caused by the abolition of the gold standard and subsequent devaluation of the dollar engineered by the Roosevelt administration.
It is important to recognize that this influx of gold was not a result of the “uncontrolled” operation of the gold standard, which had been abolished in 1933. Rather, it was the result of the deliberate and steady increase in the price at which gold was purchased by the U.S. Treasury and the Reconstruction Finance Corporation. By January 1934, the price of gold had risen from $20.67 to $35.00 per ounce, or by almost 70 percent, where it was officially pegged by the Gold Reserve Act of 1934. The Treasury was now legally mandated to maintain this devalued exchange rate between gold and the dollar by freely purchasing all the gold offered to it at this price. In effect, then, Treasury gold purchases were now economically identical to inflationary Fed open market purchases, substituting demonetized gold for government securities. Consequently, in response to this unilateral increase in the price of gold above its world price, there occurred a prodigious influx of gold into the United States — a “golden avalanche” it was called at the time — which vastly increased bank reserves. The result was an unprecedented inflation of the money supply (M2) during 1934, 1935, and 1936 at annual rates of 14 percent, 14.8 percent, and 11.4 percent, respectively.28
With respect to its influence on the supplies of bank reserves and money, the demonetized gold stock thus had been transformed into a factor “controlled” by monetary — in this case Treasury — policy. Given that the use and ownership of gold money by the public had been legally suppressed, gold was effectively demonetized and its continued purchase by the Treasury was purely a matter of discretionary monetary policy. Accordingly — and contrary to Timberlake’s assertion — when during 1937 the Treasury began to finance its purchases of gold in a manner that neutralized their effect on bank reserves, it was not engaging in deflation. The simultaneous sales of government securities to finance these purchases were simply and properly eliminating any extraneous effects of a demonetized asset on the money supply.
Even if gold were permitted to continue in its monetary function, however, Timberlake would still be wrong in criticizing the policy of neutralizing its effect on bank reserves. For under a genuine, Currency School-type gold standard, a country’s money supply would increase by exactly the amount of the gold inflow from abroad. This is not inflationary and represents precisely the proper amount by which the money supply should expand, because it is the outcome of the deliberate actions of the country’s residents who are decreasing their purchases of foreign imports and increasing their sales of exports in order to satisfy their desires for greater money holdings. This balance-of-payments mechanism is a natural part of the market economy and works continually on all levels — including the region, state, town, and even household—to efficiently adapt money supply to relative changes in money demand.
A problem arises, however, when these benign, money demand-driven gold inflows are used, as they were in the 1920s and early 1930s, as bank reserves to create unbacked notes and deposits. In this case, as F. A. Hayek has so aptly described, international gold flows will regularly cause a serious distortion of the free-market interest rate and investment pattern in the affected countries, leading to a business cycle.29 The reason is that the needed adjustment in national money supplies upward or downward now entails creating or destroying fiduciary media by expanding or contracting bank loans in defiance of the preferences of the economy’s consumers and savers. Thus, a policy of neutralizing the effect of gold flows on bank reserves in the context of a fractional-reserve banking system dominated by a central bank does not constitute a gross violation of the rules of the gold standard; to the contrary, it tends to facilitate the operation of the natural money-supply mechanism that prevails under a genuine gold standard.
Not surprisingly, in the third article of the trilogy, Timberlake also objects to the Fed’s policy of raising reserve requirements in 1936 and 1937, which was undertaken to mop up the massive amounts of excess reserves held by the banking system. Timberlake advances two criticisms against this policy. First, the policy was unnecessary because, even if all the excess reserves that existed on the eve of its implementation were subsequently fully loaned out by the banks, the inflationary potential was relatively minor. Appealing to the Banking School definition of inflation, Timberlake pronounces the 52 percent increase in the money supply that would have resulted as only mildly inflationary because the larger money supply would have exceeded the needs of trade of a fully employed economy by 5.6 percent at 1929 prices, which were about 25 percent higher than prices prevailing in June 1936.30 In plain language, Timberlake is literally defining away a potential money and price inflation of gargantuan proportions because of its perceived expedience in expanding employment and output and extricating the economy from a depression. But as Timberlake himself admits in a footnote — and as Rothbard and other Austrians have never ceased to argue — what impeded the economy’s natural and noninflationary recovery from the depression was the existence of “government programs [that] had actively worked against money price declines for ten years.”31
Growing Money Supply
In his second criticism, Timberlake contends that the increase in reserve requirements went beyond closing off a potential avenue of recovery for the economy and “turned what had been an ongoing recovery into another cyclical disaster.” But if we once again turn to Timberlake’s data we find that the money supply (M2) continued to grow, from $43.3 to $45.2 billion or by 4.4 percent, between June 30, 1936, and June 30, 1937, the year in which this policy was implemented. Even if we focus on the last six months of the period, there was hardly a wrenching deflation, as the money supply increased at an annual rate of 0.8 percent.32 Even from Timberlake’s monetarist standpoint, then, it is difficult to blame the “recession within a depression” of 1937–1938 on deflationary Fed policy.
Unfortunately Timberlake’s strained and narrow emphasis on Fed deflationism as the cause of all the woes of the 1930s causes him to ignore a plausible “Austrian” explanation of the relapse of 1937. As a result of a spurt of union activity due to the Supreme Court’s upholding of the National Labor Relations Act of 1935, money wages jumped 13.7 percent in the first three quarters of 1937. This sudden jump in the price of labor far outstripped the rise in output prices and, with labor productivity substantially unchanged, brought about a sharp decline in employment beginning in late 1937.33 The large upward spurt in excess reserves and the accompanying decrease in the money supply that we observe in Timberlake’s data between June 30, 1937, and June 30, 1938, therefore, can be explained as the result, and not the cause, of the recession.34 As business profits were squeezed by the run-up of labor costs and the economy slipped into recession, banks prudently began to contract their loans and pile up liquid reserves to protect themselves against prospective loan defaults and bank runs. To offset this uncontrolled decline of the money supply, beginning in mid-1938 the Fed (and the Treasury) once again resorted to an inflationary policy, reversing the reserve requirement increase and allowing gold inflows to once again pump up bank reserves. As a result, M2 increased by 5.9 percent, 10.1 percent, and 12.5 percent in 1938, 1939, and 1940, respectively.35
Our conclusion, then, is that the Fed’s monetary policy, except for very brief periods in 1929 and 1936–1937 when it turned mildly disinflationist, was consistently and unremittingly inflationist in the 1920s and 1930s. This inflationism was the cause of the Great Depression and one of the reasons why it was so protracted.