Editor’s introduction

This special issue of the Journal of Economic Studies has been produced to mark, as nearly as possible, the 10th anniversary of Lauchlin Currie's death. He was born in the fishing village of New Dublin in Nova Scotia, Canada, September 8, 1902 and died in Bogotá, Colombia, December 23, 1993. An economist, described as the intellectual leader of the spending wing of Franklin Delano Roosevelt's New Deal (Herbert Stein, 1969, p. 165), Currie contributed several articles and reviews to this Journal over the last two decades of his long and extraordinarily varied career (Currie, 1974, 1975, 1976a, 1979, 1981a, b, 1983, 1990). Accordingly, the Journal's editor suggested that we mark the occasion with a special issue containing some of his hitherto unpublished papers and memoranda in the field of monetary theory and policy.

As Humphrey (2003), Laidler (1993, 1999), and Steindl (1995) have explained, Currie's publications in the early 1930s presented a diagnosis of the causes of the 1929‐1932 collapse in the US economy that was substantially the same as that advanced much later by Friedman and Schwartz (1963) in their A Monetary History of the United States. In that celebrated work, Friedman and Schwartz overlooked Currie's earlier contributions. Brunner, who wrote a long foreword for the 1968 reprint of Currie's (1934a) Supply and Control of Money in the United States, gave Currie to believe that the neglect was partly due to the damage that his reputation suffered during the McCarthy era, and/or to the notion that a radical New Dealer could not also be an upholder of what became known as the conservative “monetarist” school (Sandilands, 1990, p.157)^[1]. However, in light of Laidler's reappraisal of Currie's early work, Friedman later proffered an explicit mea culpa for his earlier neglect (Laidler, 1993, p.1077n.12).

In his “restatement” of the quantity theory of money, Friedman (1956) had also contended that there was a unique Chicago oral tradition that made Chicago relatively immune to the Keynesian “virus”, and that his own restatement was in that Chicago tradition. This sparked an acrimonious debate about the nature of that tradition: whether it was unique, and whether Friedman's restatement was closer to that tradition or to the “Cambridge” tradition of Marshall, Pigou and Keynes (especially the pre‐General Theory Keynes), as Patinkin (1969, 1981), for example, maintained.

Evidence relevant to this issue was recently discovered in an hitherto almost unknown January 1932 Harvard memorandum on anti‐depression policy written by Currie et al. (2002). On seeing this document Friedman again retreated (in a letter reproduced in Laidler and Sandilands (2002)) from his earlier position on the supposed uniqueness of the Chicago tradition; see also Leeson (2003) and Smithin (2004). The 1932 memorandum indicted recent monetary, fiscal, tariff and reparations policies for causing the 1929‐1932 collapse, and called for a drastic reversal of these policies, including substantial open market bond purchases and “fiscal inflationism” – a deliberate increase in the fiscal deficit financed by monetary expansion, the latter being a necessary feature of a reactivation programme if “crowding out” effects from public expenditures were to be avoided and aggregate monetary expenditures increased.

In this the memorandum was close to the position taken at that time by Hawtrey (1929, 1931), whose assistant Currie had been during Hawtrey's year at Harvard in 1928‐1929, and close also to the position that was said by Friedman to be characteristic of a unique Chicago tradition. For example, the memorandum stated: “Our banking policy has not exerted any effective influence to check the decline in the means of payment. Instead of offsetting the decline in the demand for goods caused by the decreased rate of spending, our policy has intensified it by permitting a contraction of the volume of the means of payment.” By 1931 the deflationary forces had gone too far to be reversed through purely monetary actions:

The belated publication of this document in 2002, together with several of the hitherto unpublished memoranda on recovery policy published in the present volume, sheds new light on an important debate about the pre‐1936 intellectual antecedents to Keynes's General Theory that had already been ably discussed in Laidler's (1999) powerful book, Fabricating the Keynesian Revolution, with its apposite double entendre. Publication of the material in this volume will further contribute to this understanding, as well as to an appreciation of Lauchlin Currie's considerable influence on monetary and fiscal policy and institution‐building in the USA^[2]. As Laidler wrote to me in August 2003: “Until the profession gets it straight that the post‐war macroeconomic policy consensus had as many (more?) roots in the original work of the New Dealers as it did in JMK's efforts, its understanding of its own history will be defective.”

Currie's most notable influence in the 1930s was in drafting the 1935 Banking Act, and this volume includes the memorandum that Currie prepared for Marriner Eccles to show to President Roosevelt in November 1934 (“Desirable changes in the administration of the Federal Reserve System”, reproduced in this volume). Eccles was being considered for the post of Governor of the Federal Reserve Board but told Roosevelt that his acceptance would be conditional on the latter's support for the proposals contained in that memorandum (see Eccles, 1951, pp. 171‐3; Sandilands, 1990, pp. 62‐4).

Currie's unpublished Harvard PhD thesis on Bank Assets and Banking Theory (submitted in January 1931) was a clear antecedent both to the January 1932 memorandum on anti‐depression policies, and to his later recommendations on reform of the US banking system. The present volume reproduces three extracts of this thesis: the first and last chapters, plus a section in chapter IX on the business cycle^[3]. The thesis was begun in 1927 under the initial inspiration of Young, who at that time was himself completing an analysis of bank statistics for the USA (see Currie, 1990; Sandilands, 1999, pp. 461‐6). However, Young was then about to depart for the London School of Economics where he met his untimely death during an influenza epidemic in March 1929. Currie met him for the last time in London in the summer of 1928 while collecting statistics on the English banking system.

In the opening chapter of his thesis, Currie outlined the history of the Banking and Currency Schools debate to show the importance of the composition and control of bank assets. It reveals Currie's view that if the central bank automatically accommodates “self‐liquidating”, short‐term commercial loans (“real bills”) then, contrary to Banking School theory, this may well be inflationary and pro‐cyclical. Furthermore, commercial loans are not necessarily the most liquid or saleable, and if banks are forced to hold too large a portion of their assets in this form they may end up with excess reserves (that is, they would not be “fully loaned up”).

Later, Currie would explain why the holding of excess reserves can jeopardize the circular flow of income and frustrate the central bank's ability to control the nation's money supply via open‐market operations. Adherence to the real‐bills doctrine had, in his view, led the Federal Reserve Board in 1929 to be excessively preoccupied with banks’ security loans, hence to a tightening policy at the very moment when the real economy was moving into recession. Currie referred to this doctrine as the “commercial loan” or “needs of trade” theory of central bank policy, in which banks’ short‐term “productive” loans were prudent and non‐inflationary while their long‐term security loans were speculative, imprudent and a recipe for boom and bust in stock markets and the real economy. He blamed this theory for inducing the Federal Reserve to raise interest rates in August 1929, and for then conducting an essentially passive monetary policy during the 1929‐1932 period of mass liquidations and economic contraction (see especially Currie, 1934b).

In chapter IX of his thesis, and in line with his criticisms of the commercial loan theory, Currie endorsed, with qualifications, Hawtrey's monetary theory of the business cycle. Hawtrey focused on the “inherent instability of credit”^[4], due to the sensitivity of wholesalers and merchants (more than manufacturers) to changes in the rate of interest. The interest rate naturally falls in a depression when demand for commercial loans is low, and vice versa in the upswing. With a fall in demand for commercial loans there is a fall in output, incomes and expenditures (effective demand) until declining interest rates arrest and then reverse the fall in loan demand. In the upswing, rising incomes increase the public's demand for cash. This squeezes the banks' reserves and interest rates rise. Eventually this arrests and then reverses the upswing – in the absence of intelligent, discretionary counter‐cyclical policy on the part of the central bank.

Currie offers several criticisms of Hawtrey's monetary theory, but stresses that none are fatal. In particular, he argues that banks are able and willing to purchase “investments” (securities and bonds) when the demand for commercial loans is weak. By taking the initiative in this way they can maintain the money supply and so exert a stabilizing influence on the business cycle^[5]. He agrees with Hawtrey's critics that too much stress is put on the supply of bank deposits and too little on cash and short‐term variations in the “circuit velocity” of deposits and cash (purchasing power). Nonetheless he insists that fluctuations in the supply of purchasing power, and also in its secular trend, do exercise a profound influence on the business cycle. And if this be true then “it follows that the mechanism by which the volume of credit is expanded and contracted is also of significance.”

He goes on to state that “the study of credit movements in the business cycle involves an explanation of two problems: first, the forces affecting the reserves of member and non‐member banks, second, the means by which banks are enabled to adjust their deposits to their reserves.” The latter problem is the focus of Currie's PhD thesis on the links between the business cycle and the supply and control of various classes of bank asset. The former was more the focus of his 1934 book on the control of the supply of money. There he examined the different reserves (legally required and otherwise) held by different classes of bank during the business cycle, and the often perverse or pro‐cyclical elasticity of the money supply that had resulted from that structure.

In the concluding chapter X of his thesis Currie summarizes his main arguments and findings, including reasons for the decline of the commercial loan in banks' portfolio of assets, and the implications for monetary control and bank liquidity. (He finds, paradoxically, that commercial loans are often the least marketable, hence the least liquid of bank assets.) Referring to conditions in 1930, he complained that “[a]s long as the demand for commercial loans continues to decline the Board apparently sees no necessity of permitting the expansion of credit to take place” (pp. 243‐4). This was a recipe for pro‐cyclical monetary policy that was to deepen the on‐going depression. The Federal Reserve was too preoccupied with the composition and perceived relative productivity of bank assets, to the neglect of its primary function, that of controlling the overall volume of bank deposits:

Finally, Currie advances a number of recommendations for amending the Federal Reserve Act, especially in respect of the question of bank assets eligible for rediscount (p. 248), and of the need to relax the gold reserve requirement against notes.

Currie refined these ideas in his work at Harvard over the next three to four years. As revealed in his hitherto unpublished memoir (originally drafted in 1953) on his activities in the New Deal, this led to his being invited by Jacob Viner to join the “Freshman Brain Trust” at the US Treasury in the summer of 1934. His remit was to outline an ideal monetary system for the USA, ignoring potential political constraints (he recommended a 100 percent reserve system). This in turn led to his collaboration with Marriner Eccles and his move to the Federal Reserve in November 1934 when Eccles was appointed as the new chairman.

Brinkley (1995, p. 95) wrote that while Eccles would then come to stand at the centre of the circle of “fiscal liberals” (the spending wing of the New Deal), “he was never its intellectual leader. That role fell, for a time at least, to Lauchlin Currie, a deceptively meek young economist who was simultaneously one of the most anonymous and most influential figures of the late New Deal.” Brinkley described Currie as an iconoclastic authority on monetary policy whose severe criticisms of the US banking system in his 1934 book had brought him to the attention of the Roosevelt administration. He had complained of the “perverse elasticity” of the system because on the upswing of the business cycle the supply of money automatically tends to expand, and on the downswing to contract. It was necessary to convert the system into an effective “maladjustment‐compensating factor”.

Meltzer (2003, pp. 478‐9) also highlights Currie's influence on Federal Reserve policy after 1934, but was puzzled by the contrast between Currie's focus on monetary policy failures in his publications to 1934 and his focus on fiscal policy in his work with Eccles. He wrote:

The puzzle is resolved by a focus on the key conditions that make for effectiveness of monetary policy in the different phases of slump and recovery. To effect recovery from a slump, the Fed must first get member banks out of debt, and this the Fed signally failed to accomplish in 1929‐1932. The policy of passive acquiescence to banks' loss of deposits resulted in mass liquidations and a downward spiral into depression. An acute Hawtreyan “credit deadlock” then meant that the banks could not find sufficient credit‐worthy customers when eventually they were in a position to lend again, and so they accumulated substantial excess reserves. In such conditions monetary policy alone is, in the words of Marriner Eccles: “like pushing on a string”. The deadlock required an active fiscal policy to spend new and old money into circulation, just as Currie et al. (2002) had emphasised in their Harvard memorandum of January 1932. Equally, however, excess reserves posed a threat to effective control of a potentially inflationary recovery. Open market bond sales would not reduce the money supply if purchased out of excess reserves. Effective deflation requires that the banks be forced to borrow, and then for their repugnance to indebtedness to induce them to curtail their lending.

Thus, to strengthen the effectiveness of monetary policy in conditions of prosperity and depression, radical changes were needed. As Currie explains in his New Deal memoir, and as mentioned above, in November 1934 he drafted for Eccles a memorandum, “Desirable changes in administration of the Federal Reserve System” (reproduced in this volume; see also Meltzer, 2003, p. 467), to show President Roosevelt. The president affirmed his support, so Eccles accepted the post of chairman of the Federal Reserve, taking Currie with him as his assistant with the title of deputy director of research. Together they drafted what became, after a bitter political struggle, the 1935 Banking Act that gave them most (although not all) of what they wanted. Currie's memorandum of March 29, 1935 and speech of April 12, 1935, reproduced in this volume, explain the objectives of this legislation. The Act greatly expanded the power of the government over the banking system by moving the Federal Open Market Committee from New York to Washington, DC. It also broadened the definition of eligible assets, gave the Fed greatly enhanced powers to alter reserve requirements, and increased margin requirements for loans on securities.

An interesting part of the intellectual background to the new legislation is reflected in a letter (also included in this volume) that Currie and five of his fellow Harvard instructors sent to Roosevelt in January 1934, commending the President's controversial departure from the gold standard which “marked one of the rare occasions since the war when a government both foresaw danger and took action to avoid it.” None of the signatories was given tenure at Harvard.

While approving of Roosevelt's gold purchase plan as a means of reversing the decline in prices, the letter nonetheless warned of the danger of making direct budgetary use of the very large profits that arose from the revaluation of gold. For this would have increased member banks' reserves as an accidental by‐product of dollar devaluation and “would have called for extraordinary measures of restraint to avoid an excessive expansion of deposit money in the future.”

In the event, the huge inflow of gold from Europe (fleeing fascism) over the next few years was greatly in excess of anything that the banks could prudently lend. The substantial excess reserves that accumulated gave rise to concerns that the gathering pace of recovery might eventually engender inflationary pressures that the Federal Reserve would be relatively powerless to control. Currie discussed this in “Some aspects of the excess reserves problem”, May 18, 1936, and also in his October 26, 1936 memorandum on the implications of the recently concluded tripartite agreement on exchange rate realignments (“Recent developments in international monetary relations”). He wrote there that “the continued inflow of gold from France was proving embarrassing to us and was raising serious problems of future control in connection with excess reserves of member banks”, and he urged Eccles not to heed the pressure being put on him by Emmanuel Goldenweiser to support a return to the gold standard. Instead, he recommended that Eccles consult with Harry Dexter White, a member of the Treasury team for the Tripartite Agreement. (White's views at this time, closely paralleling those of Currie, are explained in Boughton (2002).) In a February 1937 “Memorandum on capital inflows” Currie also urged that consideration be given to introducing legislation to control short‐term (speculative or “nervous”) foreign capital in the interests of domestic stability (see also Currie, 1936).

As loan demand picked up, the excess reserves that had built up would allow the banks to accommodate the demand without restraint. This would potentially permit a multiple expansion of deposits that, together with a reversion to the historically higher average velocity of circulation (or lower demand for money as a percentage of national income), could cause monetary expenditures (MV) to expand greatly in excess of any feasible increase in real national production. The expansion in the money value of national income (Py) would then mainly be via a rise in prices (P) than in real output (y). For this reason Currie advised Eccles to raise reserve requirements in late 1936 and early 1937, but only as a precautionary measure (see, in this volume, “Would a further expansion of money be ‘injurious’?”). It was definitely not designed to be deflationary. In fact, because of the continuing inflow of gold from Europe, the doubling of reserve requirements during this period still left the banks with substantial excess reserves at the end of the process, in the spring of 1937.

However, as Currie also explains in his New Deal memoir, from about the beginning of June 1937 recovery turned rapidly into a deeply alarming recession^[6]. Real gross domestic product (GDP) would fall by 18 percent over the next 13 months (Meltzer, 2003, p. 522). The unemployment rate rose from its best 1937 figure of 12.5 percent to about 22.5 percent in the spring of 1938, including as unemployed those on emergency relief employment (Black, 2003, p. 430). At the time few people blamed the monetary measures, though there was a slight increase in bond yields, from an all‐time low of 2.46 percent in early March to 2.8 percent in early April (after which they declined again). This had angered Secretary Henry Morgenthau Jr, although it was a Treasury decision to sterilize gold inflows from December. He was worried that a rise in interest rates would increase the financial cost of the deficit, and he kept pressing Roosevelt to trim government spending and balance the budget. Eccles agreed that higher interest rates were undesirable and he supported the Federal Open Market Committee's decision in April to engage in compensatory open market purchases.

There is no evidence that the banks were suddenly denying requests for loans by business or government, or imposing stricter conditions as a result of the increased reserve requirements. Thus the almost exclusive modern focus on a supposedly inept monetary policy as the cause of the downturn in 1937 may be misplaced (see, for example, Steindl (1995, 2004), in support of Friedman and Schwartz's (1963) study). However, Meltzer (2003, pp. 521‐2) and Romer and Romer (1989, pp. 131‐2) concur with Currie by also emphasizing at least two non‐monetary forces acting to decrease output in 1937: the fiscal downturn; and the way the Wagner Act led to large inventory accumulations in anticipation of the labour market strife that did indeed occur in 1937, coinciding with an end to inventory accumulations. They also note that the behaviour of reserve holdings ran counter to Friedman and Schwartz's interpretation in that there was no discernible change in the behaviour of reserves as a fraction of deposits until December 1937, 17 months after the first increase in reserve requirements was announced and after the declines in money and industrial production were largely complete.

In a note to me, August 2, 1988, Currie admitted that probably the reserve requirements would not have been raised if the recession of 1937 had been accurately forecast. But he wrote that “this is a different matter than holding the raising responsible for the recession. For that the very sharp, even drastic, reduction in the fiscal cash deficit is the more convincing explanation of the sharp decline in the rate of growth in sales and the consequent piling up of inventories.” He also noted that “few theorists would expect an immediate impact on incomes and sales to result from the small decline in deposits that took place, especially as there is such an other more convincing explanation of the causation of the fall in aggregate demand.” These views are also spelled out and given a fuller context in a May 1938 speech to the Illinois Banking Association (“Some aspects of business and banking developments in 1936 and 1937”, in this volume). It is instructive to compare this speech with Marriner Eccles’ interpretation of the same episode (largely in terms of the non‐monetary factors at work) in chapter 3 of his autobiography, Beckoning Frontiers(1951).

A lengthy memorandum dated May 18, 1937, “An appraisal of current prospects and a tentative program” (not reproduced in this volume, but discussed in some detail in Currie's New Deal memoir) reveals that Currie was at that date relatively complacent about the restraining effects on private sector activity of a declining fiscal deficit in early 1937^[7]. Unaware just how dramatic was the decline in the fiscal contribution to spending or how great had been the accumulation of unsold inventories, he wrote: “The prospects … are for continued and modest gains in production and income unlessa net reduction of inventories occurs which, for a few months, could cause a decline in production.” However, by September of that year he was sending much more alarmist memoranda to Chairman Eccles (see below). Writing his memoir ten years before the publication of the influential Friedman and Schwartz (1963) study, it is interesting to note that Currie did not dwell very much on the role of the increased reserve requirements as a factor responsible for the 1937‐1938 recession. Instead, his focus was on the decline in the federal net contribution, cost‐push pressures, and the build‐up of inventories.

By the fall of 1937 it was clear that the economy was in sharp decline. Morgenthau infuriated Eccles (a tireless advocate of public spending) by declaring that this was proof that deficits cause recessions through their adverse effect on business confidence. He placed his faith in the driving force of private enterprise. However, the fiscal stance had been subjected to continuous scrutiny by Currie for its net income‐creating effect ever since 1934 while still at the Treasury. In collaboration with Martin Krost, a Harvard student whom Currie had brought with him to the Treasury, he developed a monthly series initially known as a “pump‐priming deficit”. These figures adjusted the government's official budget statement of revenues and expenditures to reflect the varying effectiveness of different categories on the circular flow, making allowance for those expenditure that were for currently produced goods and services and those that were merely transfers, or that merely changed savings.

In Federal Income‐Increasing Expenditures, 1933‐35, written in late 1935 or early 1936^[8], Currie and Krost reported that any similarity between the “net contribution” and the reported cash deficit was purely coincidental. The reported budget could be in balance while the net contribution was in heavy deficit. Thus there was no necessary conflict between those who wanted a balanced budget in the official sense and those who wanted the government to provide a stimulus to business: “By selecting income‐increasing types of expenditure and non‐income‐decreasing methods of raising revenue, it is conceivable that a balanced budget could be maintained and at the same time a considerable stimulus given to business.” Investment subsidies, for example, could have a powerful stimulatory effect while a tax on undistributed profits might have only a small negative effect. But there was no doubt in Currie's mind that the conditions prevailing in the mid‐1930s called for much more than a balanced expansion of taxes and spending. The size of the required deficit, whether in its cash or its “net contribution” form, was calculated according to the size of potential, full‐employment income (based on 1928 with adjustments for population and productivity growth) and the size of the leakages from that income that would need to be offset.

The 1935/1936 Currie‐Krost memorandum anticipated not only the full‐employment budget concept, but also presented a rudimentary version of the “balanced‐budget multiplier” idea. However, Currie never thought that the algebraic version that was later developed as a theorem had much relevance for policy purposes. The theorem assumed a constant marginal propensity to consume any given money supply. This implies that velocity adjusts passively to support higher incomes. However, if individuals and firms are subject to higher taxes their initial portfolios are disturbed. When their taxes are put back into the system (in practice not immediately) they may use some of this money to restore their depleted cash balances. Thus the marginal velocity (corresponding to the marginal propensity to consume) could fall in the next round and reverse the initial stimulus.

The net impact on spending would, in any case, be much smaller than if the increased government spending were financed by borrowing. So even if there were a positive balanced‐budget multiplier effect, the economic boost required in the 1930s would have called for an unrealistically massive tax‐and‐spend package. Samuelson (in Colander and Landreth, 1996, pp. 166‐7) hails the balanced‐budget theorem without addressing its realism. In this respect Currie's adherence to the period or sequence analysis of pre‐General Theory monetary theory, and his detailed studies of the demand for and supply of money in explaining the flow of aggregate expenditure was superior, for policy purposes, to Keynes's instantaneous multiplier analysis. Patinkin (1976, p. 1101) noted that Currie was one of the first economists to subject Keynes's investment multiplier to empirical test, finding that it was highly variable in the short term.

In his memoir Currie explains that because of the intense passion aroused at that time by the very word deficit, the term “pump‐priming deficit series” was soon dropped in favour of the “Federal net income‐increasing expenditures series”. It was also realised early on that in the prevailing conditions more than a one‐shot priming of the pump would be required and that deficits would probably need to be sustained for some time. According to Sweezy (1972, pp. 118‐19) the new title was “a semantic triumph of the first magnitude. It brought out the common element in all the government's fiscal operations. No one used to thinking in terms of the net contribution could advocate promoting recovery by increasing public works spending while at the same time cutting government salaries and raising tax rates.”

The income‐increasing expenditure series was to assume considerable significance in diagnosing the causes of the 1937‐1938 recession. The outlines of what Barber (1996, p. 125) has called a “domesticated Keynesian” analysis of recovery and relapse was already contained in Currie's pre‐General Theory memorandum to Eccles, April 13, 1935, entitled “Recovery” (not included in this volume). The stress there was on the importance of contra‐cyclical fiscal measures to combat a “deadlock”, with monetary policy assuming a passive role at such times. After analysing the prospects in various fields, Currie concluded that:

As economic recovery faltered in mid‐1937 it became evident that “pump‐priming operations” had not been sufficiently vigorous or steady. Fiscal policy was now operating in a perverse direction. In February, Currie had already condemned the 1937 Social Security Act because of the deflationary implications of building up a large reserve fund, especially when in 1937 there was nothing to replace the large pay‐out of veterans' bonuses in 1936 (passed by Congress over the president's veto).

Telser (2003, p. 240) has emphasised the importance of the veterans' bonus payments (mostly in 1936) on the buoyancy of recovery in that year, even though it was financed by borrowing from the public so that it did not change the money supply. He states that the fiscal deficit in June 1937 was less than a quarter of its June 1936 level (Telser, 2003, p. 238), and so it had a correspondingly smaller impact on the economy in 1937. On Currie's measure of the “net federal contribution”, the fall was even greater ($101 million in June 1937 compared with $543 million in June 1936)^[9]. This may be partly explained by the differing spending propensity of the veterans compared to those who financed their bonuses.

By the autumn he was sending increasingly urgent memoranda. One of these “Comments on business prospects” (September 28, 1937), begins on a curiously upbeat note: “The national income for the year as a whole should make a very gratifying comparison with the previous year.” However, after listing some of the positive results for the first six months of the year, he noted that “[a]s indicated in the recent study by [Arthur] Gayer and [Martin] Krost it is expected that the net contribution of the Federal Government to total community expenditures will decline in the fiscal year 1938 by some figure in the range from $2.5 billion to $3.8 billion. This will be a drastic reduction.” He also emphasized that there had been a rapid advance in building costs (hourly wages in the construction industry had increased by 16 percent in little over a year, and building materials prices by 13 percent) relative to the increase in rents, and residential contracts awarded had been declining since June.

In “The decline in the federal contribution to the growth in community expenditures” (October 19, 1937), he showed that in the three years 1934‐1936 the net contribution had been $3.2 billion, $3.1 billion, and $4.0 billion. These were sizable fractions of the growth of national income in those same years: $7.8 billion, $5.4 billion, and $8.8 billion respectively. In the eight months from February to September 1937, the Currie‐Krost series showed that the net contribution had fallen to only $573 million ($68 million a month and still falling – it was estimated at only $37 million in September) compared to $2,789 million ($348 million a month) in the same period of 1936. He warned that the government's contribution to buying power, already insufficient to offset the slowdown in private expenditures, may well turn negative in the near future.

On a Keynesian interpretation of the downturn in 1937‐1938, this fiscal reversal was a crucial causal factor. By comparison, variations in the degree of excess liquidity in the banks were of secondary importance. By cutting the federal deficit there was a fall in the supply of safe earning assets that the banks had previously relied on^[10]. They could not easily or quickly replace them with a corresponding increase in private sector lending, for the decline in government spending and the increase in tax and social security receipts were themselves depressing demand for private sector output. This naturally restrained private sector loan demand. These factors, rather than the raising of reserve requirements, may account for the diminution of demand deposits from mid‐1937 and the continued high level of excess reserves.

Telser (2001) also disputes Friedman and Schwartz's (1963) view that the 1937‐1938 recession was caused by the raising of reserve requirements. He shows that there was no decline in bank lending to the private sector until the end of the first quarter of 1938, hence higher reserve requirements could not explain the decline of business. The decline in banks' earning assets was entirely in their holdings of government bonds which were sold in order to meet reserve requirements. The changed composition of bank assets (as well as their decline) differed radically from the experience of the previous recessions of 1920‐1922 and 1929‐1933. But if this meant that the private sector was not starved of loans, the question remains: “What did cause the 1937‐1938 recession in private business?”

Telser does not explicitly address this. Currie's explanation, however, was that the fiscal tightening of 1937 led to a much smaller supply of new and relatively riskless bonds for banks to purchase. Thus, even without an increase in reserve requirements they would have been liquidating maturing bonds. Also, he insisted that they were inclined to sell bonds anyway because of profit‐taking at the apparent end of a three‐year bull run for bond prices that occurred in January 1936 even before the announcement of the rise in reserve requirements. In his May 1938 speech to the Illinois Banking Association he concluded:

If banks had continued to buy bonds on secondary markets this would have increased the riskiness of their asset portfolio by further depressing interest rates. If they had maintained their assets by making more fresh loans to the private sector in place of loans to government, this would also have increased their risk, absent a strong increase in loan demand by credit‐worthy customers. There was some increase in private sector loans throughout 1937 but insufficient to compensate for the public sector's decreased demand, especially as private profit prospects were dented by the decline in the government's “net contribution”. Thus if the Federal Reserve had not raised reserve requirements there would probably have been further accumulations of excess reserves.

Nonetheless, the actual policy of increasing reserve requirements under these conditions could not have helped matters, and would partly explain why banks sold bonds to maintain their liquidity. Again, however, Currie gave greater weight, in his May 1938 Illinois address, to banks' desire to avoid capital‐value losses at the end of the bond market's bull run, with the decline “initiated by municipal and Federal bonds in January [1937] before action with reference to excess reserves was announced”^[11]. He insisted that action on reserves would not have worsened the recession whose causes lay elsewhere – in the smaller deficit plus other non‐monetary factors, notably the exceptional inventory accumulations in late1936 that would be worked off a few months later.

Another effort to explain the importance of the Federal Government's net contribution to buying power as an offset to, as well as an explanation for, the increased size of idle bank balances held by different groups (consumers, business, public bodies, foreigners, and financial institutions), was presented to the American Statistical Association in Atlantic City in December 1937 (“The economic distribution of demand deposits”; included in this volume). This was an interesting example of his efforts to go beyond the mechanical measurement of velocity by concentrating instead on the reasons why different groups increase or decrease their demand for money as a proportion of their incomes, and the relative weights that attach to each group at different phases of the business cycle.

His analysis reinforced his view that the driving force of recovery and relapse in 1936‐1937 was the size of the government's net contribution: “The continuation of activity‐stimulating expenditures of the Government was necessary because the initial impetus of Government spending quickly lost its momentum in the conditions prevailing in this period. A portion of the receipts of both business and consumers remained unspent.” However, he still anticipated that as and when recovery began anew, the monetary requirements of business and consumers could be met without necessarily entailing a rise in interest rates or the creation of new money. Indeed, it still implied that the Federal Reserve needed effective power to control an excessive growth of money in the upswing. But with excess bank reserves, money and velocity could together be dangerously endogenous to business conditions.

Steindl (2004, p. 66) has written that in his analysis of recovery Currie had abandoned the quantity‐theoretic analysis that he had applied to the contraction of 1929‐1933: “He therefore did not see the recovery as the product of an increasing stock of money. For him, the quantity of money was now an endogenous variable, subject to the needs of business as it sought to borrow, thereby affecting deposits, money, and excess reserves.” The quantity theory's identity (MV=Py) states that if the velocity of circulation is constant then the value of national income (Py) will increase in line with M. But in the unusual conditions of the mid‐1930s Currie feared that both velocity and money would be excessively pro‐cyclical in any firm recovery, since business and consumers would be able and willing to draw down their substantial idle balances^[12]. Steindl's belief, based on Currie's published work, that he showed little interest in the money supply after 1934, is not borne out by his unpublished memoranda reproduced in this volume, though it is certainly true that he continued to believe, as he had done ever since the initial failure of the Fed to avert economic collapse in 1929‐1932, that monetary policy had now become relatively powerless as a recovery measure except in conjunction with a vigorously expansionary fiscal stance. Its role would remain a subsidiary or complementary one until full recovery had been firmly established.

In another sober assessment of the 1936‐1937 episode, in his “100% reserve plan” memorandum (August 12, 1938; included in this volume), Currie also lamented the diffusion of authority and responsibility for monetary policy between the Treasury and the Federal Reserve Board, with the Treasury having sterilized gold at the same time as the Board was raising reserve requirements. This had produced a combined effect greater than would have been desired by a unified authority with enhanced powers. He also blamed the local character of American banking as a main factor behind its terrible loss record, its exceptional emphasis on liquidity, and the wide swings in demand deposits. Nationwide branch banking (as in England and Canada) would go a long way to alleviating these defects, but this had been prevented by America's deep‐seated distrust of financial concentration. However, he renewed his call for 100 percent reserves against demand deposits (with none against time deposits) as another way to break the destabilizing link between bank lending and media of exchange (the demand liabilities of the banks). Cash drains would also lose their destabilizing effect^[13].

Even as he was calling for an increase in government spending and reduced taxes, Currie was suggesting, in an untitled six‐page memorandum dated October 21, 1937 (not included in this volume), that it be financed not by sale of bonds (which would require that the Treasury compete for funds with the private sector) but rather by desterilizing some of the Treasury's inactive gold account. However, this can only be done, he wrote, “if the Board of Governors is given power to absorb the additional reserves. Otherwise there would be grave danger of inflation. The Board should be given authority to impose increased reserve requirements on foreign balances and on banks which are now non‐member banks. Authority to modify or further increase existing reserve requirements for member banks may also become necessary.” If the 100 percent reserve plan was a non‐starter, then he nonetheless appeared to believe that a 100 percent marginal reserve plan could be implemented.

On November 8, 1937, WPA administrator Harry Hopkins and his economic adviser Leon Henderson, together with Currie and Isador Lubin, Commissioner of Labour Statistics, met with the president in an unprecedented four‐hour session (see Lash, 1988, pp. 317‐27). The New York Post reported the next day that “the four advisers minced no words in giving Roosevelt a hard‐boiled review of economic conditions and with equal bluntness and vigor they told him that a disastrous recession can only be averted by a resumption of big‐scale Government spending.” The group laid a report before the president that showed that “in August, for the first time since 1931, the government took more out of the income stream than it poured back in … If the Government takes taxes away from workers or corporations and uses these in bookkeeping items, such as old age reserve accounts, gold purchases, debt retirement, etc., and the amount exceeds what is paid for men and materials, then there is a deficit. That is what is happening now.”

Here they were using the term “deficit” to refer to a deficit of overall spending, not the budget deficit. The deficit in the income stream had to be reduced by increasing the federal contribution; that is, by increasing the budget deficit. But in a speech the very next day (November 10) Secretary Morgenthau declared that the latter deficit was excessive. A balanced budget was needed to restore business confidence.

As Stein (1969, Ch. 6) put it, the Keynesians and the budget‐balancers were now locked in a furious “struggle for the soul of FDR”. The report that the Keynesians placed before the president stated that if the government continued to take out more than it puts in then “(1) prices will not adjust quickly enough, (2) budget balancing will be pursued too far and deflation will result, (3) unemployment will increase, (4) buying power will be impaired, and (5) things will get out of hand.” The large increase in production in 1936 should have led to a vigorous increase in retail business in 1937. Instead, a combination of cost‐induced (as distinct from demand‐induced) price increases and cuts in government spending meant that the large increase in production could not be taken off the market because purchasing power was inadequate. Cost advances in the key construction sector were also highly damaging and needed to be offset by reduced financing charges. The report continued:

In fact the economy was in a tail spin. In a speech to Congress a few days after his “Keynesian” seminar, Roosevelt asked: “What does the country ultimately gain if we encourage businessmen to enlarge the capacity of American industry to produce unless we see to it that the income of our working population actually expands sufficiently to create markets to absorb the increased production.” But in practice Roosevelt initially sided with Morgenthau. In Brinkley's (1995, p.28) terse words, disaster followed. Not until April 14, 1938, after the worst period of his long tenure in the White House and after a strong letter from Keynes in February, did Roosevelt at last ask Congress (over the continuing objections of the Secretary of the Treasury) for more than $3 billion of spending or lending in the immediate future for relief, public works, housing and assistance to state and local governments (Barber (1996, p. 114(; see also Black's (2003, pp. 428‐36) much acclaimed biography of Roosevelt).

The New Deal of course was about much more than the size of public spending. However, in the absence of macroeconomic balance relatively little could be expected of microeconomic reforms. Gardiner Means continued to insist that laissez‐faire was played out and that detailed industrial planning was called for to eradicate the malevolent influence of administered prices and the output‐suppressing propensities of producers with market power (Barber, 1996, p. 126; Lee, 1990). By 1938‐1939, however, the stress was on spending first, structural reform second. The spenders thought that monopoly was as much the consequence as the cause of depression. Expansion of the market, domestic and foreign, would offer opportunities for greater competition from new firms and products.

On April 29, 1938 President Roosevelt delivered a “Monopoly message” to Congress in which he proposed an appropriation of $500,000 to fund an exhaustive investigation into the concentration of economic power. The resultant Temporary National Economic Committee (TNEC) was to generate 31 volumes of testimony over the next three years. Currie thereupon persuaded Leon Henderson and Jerome Frank at the Securities and Exchange Commission (SEC), the main instigators of the TNEC, to include the study of macroeconomic policy as well as the study of monopoly and industrial concentration. Stein (1969, p. 168; Brinkley, 1995, pp. 128‐36) observed that the TNEC hearings turned out to be mainly a showcase for Keynesian economics, with Currie and Hansen the star witnesses, having teamed up as “Mr Inside and Mr Outside” (Tobin, 1976, p. 33) to present complementary presentations in May 1939 of the theoretical and empirical case for compensatory fiscal policy (see Currie, 1939). Shortly after this, in July, President Roosevelt (apparently on the recommendation of Irving Fisher^[14]) made Currie even more of an insider by bringing him to the White House as his administrative assistant for economic affairs.

Currie retained this position as the White House economic adviser until after FDR's death in 1945 (interspersed with responsibility for administering the lend‐lease programme to China and the Foreign Economic Administration). In his work during 1939 and 1940 he continued to press for expansionary spending programmes to deal with the persisting tragedy of unemployment and unused capacity. The subjects he dealt with ranged widely, from plans to boost power supplies, transport and military equipment in preparation for “defence” (war), social security legislation, banking matters, the annual budget, farm security, housing, a national programme for youth, foreign economic relations, and political matters in the run‐up to the 1940 presidential elections. His March 18, 1940 “Memorandum on full employment” and his December 2, 1940 memorandum, “Expansion possibilities of our system” (both included in this volume, and also discussed in Currie's 1953 Memoir) have the most distinctively Keynesian flavour of all the papers in this volume, containing as they do explicit recommendations on, for example, how to boost the propensity to consume via a more progressive tax system. Similarly, in a fiscal policy memorandum dated June 4, 1940 Currie insisted that “a fundamental fact at the moment is that America possesses an overall great excess of unutilised material and human productive resources. Hence, the financing of national defence should be directed at taking up this slack rather than toward the diversion of resources from the making of consumer goods to the manufacture of armaments. This latter course is proper only when the economic machine is working at full capacity.”

The lasting legacy of the theoretical, empirical, and practical experience of the depression and war years was the February 1946 Employment Act and the creation of the Council of Economic Advisers. Its passage through Congress was stormy, and the original bill that Alvin Hansen drafted in August 1944 was much watered down. Nevertheless, a statute that affirmed governmental responsibility for “maximum employment, production and purchasing power” was a significant advance over the much more limited mandate for government that, for example, was preferred by Irving Fisher and the Chicago School with their rules‐based price stability goal for monetary policy, and with fiscal policy aimed at low‐level balanced budgets. The war itself accustomed people to higher and more progressive rates of taxation and government spending, and these were only partially retrenched in peacetime. This introduced a much greater degree of built‐in stability by effectively increasing the marginal savings rate at the full employment level of income and expenditure.

Shortly after Roosevelt's death in April 1945 Currie resigned from the US Government and set up in business as a private economic consultant. In 1949 he was engaged by the World Bank to direct a comprehensive survey of economic conditions in Colombia prior to the granting of a major loan package (Currie, 1950). The Colombian Government subsequently contracted him to advise on the accompanying policies and institutional reforms that were needed to ensure that the loans might have maximum development impact. Naturally, one of the key issues that Currie focussed on concerned monetary and exchange rate policies and the structure of the banking system. In 1949 Colombia was predominantly agricultural (with coffee the main export crop) and development demanded major shifts in the structure of production toward urban‐based industry and services. Colombia's central bank, the Banco de la República, has traditionally seen its role as twofold:

• 1.

  Accelerating development by selectively advancing “productive credits” to agriculture, industry and government.

• 2.

  Control of the nation's money supply (as well as its role as supervisor of the commercial banking system).

These two roles have the potential for great conflict, and Currie struggled over the years to persuade politicians that the central bank's main role should be to ensure price stability via monetary discipline, while at the same time developing separate institutions that specialise in financial intermediation between savers and borrowers with minimum recourse to inflationary finance.

There are many similarities between Currie's critique of central bank policy and institutions in Colombia and his discussion in the 1930s of the policy mistakes that arose from loose definitions of “money” and “credit”, the fallacies of the real bills doctrine, and the need to reform the powers and responsibilities of the Fed. In Colombia he insisted that a central bank should not be a development bank and that it should be ever conscious of the dangers of conducting monetary policy via forced investments imposed on the commercial banking system (with its inflationary bias) rather than via strict control over (frozen) reserve requirements as a means of controlling the circular flow of the nation's monetary incomes and expenditures.

In practice, Colombia has relied heavily on the printing press for finance with the inevitable consequence of chronically high and variable rates of inflation. This plays havoc with the allocation of resources as between long‐ and short‐term projects. Currie's main concern was that inflation severely suppresses the latent or potential demand for long‐term mortgage finance, and hence has greatly inhibited the urbanisation process. Without adequate housing finance the mobility of labor out of low‐paid, low‐productivity peasant agriculture into higher‐productivity urban work has been seriously affected. This has damaged both the efficient allocation of resources and has also prevented a more rapid alleviation of poverty and improvement in the distribution of income.

As the author of a celebrated national plan, the “Plan of the four strategies, 1972‐74”, one of Currie's major achievements in Colombia was to conceive and guide the implementation of a new housing finance system based on the indexation of both its assets and liabilities^[15]. The result was a major boost to the capture of non‐inflationary savings on the one hand, and the simultaneous increase in demand for them by households that previously had to pay very high nominal rates of interest on their mortgages. With index‐linked loans they could now pay a more modest nominal interest rate (but positive in real terms, to make it attractive to savers) while spreading the real amortisation of the capital more evenly over the life of the mortgage. This system has played, and continues to play, a major role in Colombia's modernisation. Currie published widely on this theme (for example, Currie, 1974, 1976b, 1981a, b, 1997).

Meanwhile, however, money and inflation in Colombia have seldom been under firm control. An understanding of the nature of inflation and its control requires not only an understanding of the levers of monetary control on the supply side, but also of the factors that influence the demand for money. In the 1930s at the Federal Reserve Currie had conducted a pioneering “Large deposits study” (May 1936), with survey results that indicated that a very large proportion of the nation's money supply was accounted for by a very small proportion of the total number of the banks' current account customers (mainly large corporations or public bodies). As noted above, he presented some of the results at the meetings of the American Statistical Association in December 1937, as “The economic distribution of demand deposits” (also see “The behaviour of deposits”, April 24, 1938, included in this volume). He showed that an understanding of the velocity of circulation (or the inverse of the percentage of national income held as money balances) therefore depended crucially on the behaviour and motivations of the people managing these large accounts. In the late 1980s Currie was able to obtain similar statistics on the ownership of Colombia's money supply. Again he found that it was heavily concentrated in the hands of a few owners of very large current account balances.

He thus developed a new hypothesis on the demand for money based on the cost to the banks of administering these large and very active current account balances, and the potential earning of the banks from the lending of these liabilities, which in turn depended on the going rate of interest. The banks will accordingly vary the size of the minimum balances that their customers are obliged to hold and/or the charges they will impose for every cheque that is cleared if the minimum balance is not maintained. So far as I am aware this is a quite novel approach to an understanding of secular and cyclical variations in the demand for money as influenced by changes in bank costs (as affected by technical innovation in the banking industry^[16]) and the interest rate (as affected by the business cycle). The theory differs significantly from the traditional textbook discussions of the transactions, precautionary, and speculative demands for money. Currie argues that the chief weakness of the traditional theories is that they focus only on the costs to the holders of money and not on the suppliers of money, with the bulk of it supplied by commercial banks. He also shows how this reformulation helps in settling the important question of the most useful definition of money, both in terms of clarity of thought, and therefore also for purposes of more effective monetary policy:

Currie completed various drafts on this subject shortly before his death and his November 1992 manuscript is published in this volume for the first time. It may be taken as the culmination of his deep thinking on monetary theory and his vast experience in the conduct of monetary policy over six decades.

As a sequel to this introduction, I give a biographical sketch that reveals the richness of Currie's overall career as an academic economist and practising economic adviser at the highest levels in the United States and Colombia. Another version of this sketch may be found in the American National Biography On‐line at http://www.anb.org

Note: The Editor thanks Elizabeth Currie (of the World Bank) for permission to publish this selection of her father's hitherto unpublished papers. The bulk of Currie's papers, published and unpublished, is held by Duke University (NC) in their Rare Book, Manuscript, and Special Collections Library (http://scriptorium.lib.duke.edu).