♦
This chapter examines the financial links between an international economic system and domestic economies. The world depression of the late 1920s and early 1930s was an era of budgetary orthodoxy. In most countries, governments reacted to declining prices and economic activity with attempts to balance budgets, with a deflationary result: the reduction in aggregate demand intensified the process of decline, and in this way the public economy played a major role in the transmission of deflation (and thus depression) across nations.
Governments felt powerless in stabilizing expectations because of three circumstances. First, financial and banking systems were volatile and vulnerable to panic. The less stable the banking system, the greater the implicit fiscal liability, as governments were expected to tackle the damage inflicted on the economy by failed banks, but also the greater the helplessness of the government. Second, there existed no agreement or consensus about what governments should do domestically. What consensus might have existed previously evaporated during the depression. Third, there was an increasing consensus that the true function of government was an international one, to externalize adjustment costs (less politely, to make the foreigner pay).
At the outset of the story, countries looked very different from one another in terms of economic potentials and problems. But after the financial panic had set in, each country looked unhappy in a pretty standard way. Fiscal and financial crises reinforced each other: fiscal difficulties led to capital flight, and the withdrawal of capital weakened banks and created a potential or actual fiscal burden. Banking problems thus led to fiscal problems, because the cost of taking over bad banks strained the budget. But budget imbalances were interpreted by investors, foreign and domestic, as meaning that there were limits to the government’s ability realistically to offer support for banks, and that it was therefore time to get out.
The story of differing national origins of a crisis that would eventually become a regional and then a global one is familiar from analyses of the crises of the 1990s, in which “twin crises”—banking and financial—reinforced each other but developed along particular paths.1 Thus Thailand in 1997 had a problem primarily because of the unsoundness of its banking system. Malaysia had experienced an asset boom (or bubble), along with a surge of public-sector investment, and although the bubble had burst in 1993, the banks camouflaged their bad loans. Korea had a very high foreign currency debt and insolvent conglomerates (chaebols), with a wave of corporate bankruptcies beginning in January 1997; but the foreign debt crisis occurred nearly twelve months later, when the threat that foreign short-term bank debt could not be rolled over set off a general flight. These different experiences of crisis and the threat of crisis rested on two common bases: a surge in foreign lending in the first half of the 1990s, including loans to banks that had a weak capital structure; and an approach to exchange-rate policy in which a fixed rate encouraged capital inflows taking advantage of interest-rate differentials. In these modern crises, the monetary authorities—like their interwar predecessors—faced impossible choices. They could attempt to defend their currency by interest-rate increases, but this course risked damaging the domestic credit structure, as borrowing would become impossibly expensive and lead to a wave of domestic bankruptcies. Alternately, they could abandon foreign-exchange pegs and let their currencies slide. This approach would damage the banks and corporations that had borrowed cheaply in foreign currency and then lent in domestic currency. A devaluation would be likely to lead to further currency declines as the effect on the corporate sector became apparent, and massive capital flight would follow. These modern policy dilemmas have precise analogues in the unpleasant character of policy choices available at the beginning of the 1930s.
The hopelessness of the situation in the interwar period not surprisingly led to a way of thinking about depressions that emphasized irrational or psychological factors. Such interpretations, which have been popular at regular intervals (often coinciding with depressions), depend on a belief that markets are driven by psychological and nonrational calculations.
The mechanism of financial panic played a crucial role in linking the depressions in the capital-importing world (the “emerging markets” of the 1920s) and the capital exporters (the advanced industrial countries, in particular Britain, the United States, and France). Modern writing on the transmission of crises and on early-warning signals of potential banking problems emphasizes that an important and reliable sign of an imminent financial problem is the extent of banks’ overseas short-term liabilities. Banks are especially vulnerable in the aftermath of inflations, hyperinflations, and financial liberalizations that may attract large sums of foreign capital. In their function as capital exporters, the large industrial countries also received extensive short-term deposits from the countries to which they lent. This exposure to risk played a major role in the development of a second stage of the crisis.
The following sections examine why international action and coordination of policy responses against the forces of depression and deflation were so ineffective, why crises converged in the emerging markets, and how panic spread to the industrial world and the dominoes began to topple, one by one.
The international financial system of the interwar period was more ordered and regulated than the nostalgically celebrated prewar gold standard. Whereas the nineteenth-century gold standard had evolved by accident, because in the 1870s Germany and the United States chose a gold-based currency, and thus created a bandwagon effect,2 the restoration of gold in the 1920s corresponded to a plan. This plan, elaborated most clearly in 1922 at the Genoa international monetary conference, involved: gold convertibility, the establishment of independent central banks, the disciplining of fiscal policy, conditional financial assistance for countries on the margins of the system, and the continued cooperation of central banks in the management of the system.
Gold Convertibility
Peter Temin’s survey of the depression, and Barry Eichengreen’s monumental history of the “golden fetters” that limited policy, attribute the severity of the Great Depression to the policy adjustments required in the aftermath of the adoption of a restored gold (or gold-exchange) standard.3 The question Temin raises is a variety of the old “confidence” chestnut: the gold standard was seen as an essential part of confidence, but it also created a framework in which confidence mattered more. Would it have been more reassuring to have a system less dependent on psychology, in which flexible (or floating or free) exchange rates increased the risks for speculators playing the game of “confidence”? John Kenneth Galbraith had already identified the false pursuit of the chimera of confidence as a crucial weakness of interwar economics, “the dangerous cliché that in the financial world everything depends on confidence.”4
Such was in fact the immediate postwar situation, from which Europe struggled painfully to emerge after the early 1920s. It proved inflationary, and unconducive to long-term international flows of productive investment.
In the early 1920s, as inflation ravaged large parts of Europe and violent revolution, civil war, and even new international conflicts seemed imminent, “experts” and politicians met in conference at Genoa. They set out to work out a way of restoring the old order: bringing domestic and international peace by the expansion of trade and the stabilization of currencies. At the heart of the new version of how to create postwar stability was the resurrection of a functioning international payments system, for without it, hopes for greater international trade would be as futile as those for domestic stability. The experts at Genoa intended the gold-exchange standard to be as little deflationary and as painlessly inflationary as possible, since in the early 1920s even the dullest analysts saw inflation as a way—though not a costless one—of reducing social tensions.
A subsequent paper produced by the Bank for International Settlements described the system devised by the Genoa experts as having an “undeniable tendency towards credit expansion,” since the international stock of currency reserves was much increased by the inclusion of key currencies (sterling and the dollar) as well as metallic money.5 In fact some countries, notably Japan, had already maintained an exchange standard before the war. This step had represented a considerable economy, in that reserves could be held in the form of commercial or treasury bills, normally in London, rather than as unproductive and non-interest-bearing gold.
The experts who devised schemes such as that of Genoa were supposed to offer neutral, apolitical, or at best depoliticized advice and solutions. A decade later the Swedish economist Per Jacobsson gave a fine but disenchanted definition of what this expertise involved: an expert was “a man who can express the opinion of his Government in technical terms.”6 The postwar inflations that they were expected to solve had a technical and immediate cause—unbalanced budgets—but also a much broader political and social background. The experts realized this well. One of the most intelligent observers, the director of the League of Nations Economic and Financial Section, Sir Arthur Salter, told the U.S. Senate’s Commission of Gold and Silver Enquiry: “There is still sometimes too great a disposition to regard inflation as merely a financial vice, a sort of post war Finance Minister’s drug habit. It is too little recognized that it was in many cases the only practicable method of avoiding social collapse in the conditions left immediately after the War. Inflation is, in my view, the practically inevitable complement of war and post war [domestic] loans after these passed a certain proportion of the national income or annual taxable capacity of a country.”7
The experts’ solutions aimed at creating an automatic mechanism that could control the instincts of political parties and pressure groups to push for the continual expansion of state expenditure even while demanding reduced taxation. Budget balancing, imposed through external constraints, meant a limitation on the political process and on national sovereignty. It was in the latter sense that fiscal orthodoxy formed a part of the creation of the new international order of the 1920s.
The pleasing ambiguity of the gold-exchange standard, which helped to secure its political acceptability, was that it provided at the same time a restraint and an element of international order, and also that no one could tell quite how restraining it would turn out to be. Its charm lay precisely in that it seemed to offer a dry path between the two ditches of inflation and deflation.
Independent Central Banks
The doctrine of central banking held that monetary authorities should be independent of governments so that they would not need to respond to political pressures. They should run monetary policy in accordance, not with domestic priorities, but rather with the requirements of the international system.
The most forthright exponent of this theory, the long-serving governor of the Bank of England, Montagu Norman, made quite explicit the political role that the central bank needed to play:
Central Banking is young and experimental and has no tradition: it may last and may develop, or its usefulness, to fill a short post-war need and no more, may soon come to an end. On the one hand its sphere is limited by the qualification that no Central Bank can be greater than its own State—the creature greater than the creator. On the other hand, a Central Bank should acquire by external help (as in some ex-enemy countries) or by internal recognition (as in France) a certain freedom or independence within, and perhaps without, its own State. By such means alone can real co-operation be made possible. I cannot define the position thus acquired but it should surely permit a Bank to “nag” its own Government even in public and to decide on questions on other than political grounds.8
Central banks were to be established before a country stabilized on gold in order to prepare the institutional ground. The Brussels Conference made the same point as Norman when it concluded in 1920, “Banks and especially Banks of Issue should be free from political pressure and should be conducted only on lines of prudent finance.”9 The constitutions of the new banks were thus framed with clear political objectives. Norman made the point quite explicit: “It seems evident that the limitations imposed on new or reorganised [Central] banks during the last few years arise more from the fear and mistrust of political interference than from the needs of Central Banking as such.”10
But it was not just vis-à-vis their governments that central banks needed independence: they were also threatened by the claims of their commercial banking systems. As continental central banks fueled inflation by cheap rediscounting of bills (the German Reichsbank raised its rate only in 1922, long after the inflation had become hyperinflation), Norman declared: “A Central Bank should protect its own traders from the rapacity of other banks in his own country.”11
The new central banks of the interwar years included those associated with currency stabilization schemes in Austria, Hungary, and Germany; but the principle was extended throughout the world. Chile had a new bank in 1926, Argentina in 1936, and in Brazil commission after commission staffed with British or American advisers recommended the introduction of this institution. Canada established its bank in 1935.
The central bank governors saw themselves as members of a club, engaging in friendly and intimate relationship with one another. “You are a dear queer old duck and one of my duties seems to be to lecture you now and then,” wrote Benjamin Strong of the Federal Reserve Bank of New York to Montagu Norman.12 In particular Strong, Norman, and their German colleague Hjalmar Horace Greeley Schacht shared a similarity of outlook and behavior. After Strong’s death in 1928 from complications following from tuberculosis, his successor, George Harrison, never recaptured the imagination or loyalty of his European colleagues. Five years after Strong’s departure, Schacht wrote to Norman: “I feel most strongly that, after the death of our American friend, you and I are the only two men who understood what had to be achieved.”13 France, which stabilized relatively late, remained on the edge of the charmed circle of bankers.
The Disciplining of Fiscal Policy
The immediate purpose of the adoption of the gold-exchange standard and the strengthening of central banks was control of fiscal policy. It was an ambitious effort, because the whole foundation of nineteenth-century prudent public finance had been destroyed by the Great War. No country could finance a total twentieth-century war through taxation. In every state, the major source of war finance had been borrowing—either through bond issues or, as governments became increasingly desperate, through short-term issuing of treasury bills. As a consequence, every state emerged with a high public debt: the Russian empire’s increased by a factor of four, Italy’s and France’s by five, Germany’s by eight, Britain’s by eleven, and that of the United States (where there had been only a very small public debt before the war) by nineteen.14 In addition to the cost of servicing this debt, the legacy of war inevitably brought new expenditures. The most obvious were the pensions for war widows and for those crippled in battle.
After the war there appeared to be a stark choice: either to revert to fiscal rigor or to wipe out the national debt with inflation. The first required a conviction that sacrifice could be imposed without provoking revolution. Measures such as the British “Geddes Axe,” the report of 1922 by a committee on public expenditure under Sir Eric Geddes, recommended military cuts, reductions in educational spending, and the abolition of five government departments. Continental Europe, with stronger left wing and revolutionary movements, largely felt unable or unwilling to take such action, and opted for inflation instead. But this too had an immense—and rising—social and political cost. Alone in central Europe, the Czechoslovak finance minister Alois Rašin tried to found his new state on the basis of a sound currency. As surrounding states collapsed in inflationary disasters, capital flooded into Czechoslovakia and drove up the exchange rate. The Czech crown rose in value against the stable Swiss franc from 5 centimes to 20 centimes in September 1922. Such a real appreciation did great damage, and Rašin became massively unpopular. In January 1923 he was assassinated.15
Given such political constraints, an externally imposed stabilization—with the gold standard as a nominal anchor—looked like the only way of explaining to a domestic audience why fiscal expansion had to stop.
Conditional Financial Assistance
The first stabilizations that reflected this new “expert” and “apolitical” order occurred in Austria and Hungary after the initial breakdown of the Paris peace treaties.16 Both countries were in a hopeless position, and it soon became apparent that they could not pay war reparations. In May 1921 the Allied governments had wound up the Austrian section of the Reparations Commission. The next year the League of Nations devised a stabilization scheme for Austria involving the floating of a foreign loan, the issue of a new currency (schilling), the creation of an independent central bank, and above all the imposition of financial control by a League commissioner, the Dutchman Dr. Alfred Zimmermann, who would control the distribution of League funds. Many public revenues would go directly to the commissioner rather than to the Austrian government. The immediate price to be paid—to show that austerity was being implemented—involved the dismissal of 100,000 Austrian civil servants.
Hungary had a similar scheme, devised in December 1923, again with a bank of issue, a new pengö currency, a 7.5 percent League loan of 250 million gold crowns ($50 million), a League commissioner (Jeremiah Smith), a British adviser to the national bank, and an austerity program agreed in consultation with the League. By 1925, 25,000 Hungarian government jobs had been suppressed.17
The economic situation of Hungary was slightly better than that of Austria, but the political position was worse, with a revolutionary Communist dictatorship in 1919 whose rule had been ended only by an invasion of the Rumanian army. In its place came a right-wing authoritarian regime, which the League’s spokesman termed “strong, stable, and drastic.” The international politics of the Hungarian stabilization were also more difficult. Before the package was complete, the League needed to consult not only with the great powers but also with Hungary’s angry and in part unstable neighbors, Rumania, Yugoslavia, and Czechoslovakia. Clause VI of the scheme devised by the League Financial Committee stipulated “satisfactory political relations between Hungary and her neighbors.”18
These schemes for Austria and Hungary, which indeed halted inflation and transferred the blame for this costly operation from the national government to the international organization, provided a model for later stabilizations.19 But so great was the political humiliation that no other government was willing to go to the League. Subsequent stabilizations, such as that of Germany, were worked out by the great powers or—increasingly—by the private capital markets. The League complained bitterly that the availability of bank money was undermining its attempt to restore general principles of fiscal rectitude. “If American money is going to be made available in this way,” the senior League financial official wrote, “we must either expect to have come pretty well to the end of our financial reconstruction or we must have some effective way of bringing American banking into our organisation.”20 The whole process was similar to that of modern developing countries running away from the allegedly harsh terms of the IMF and finding private-sector lenders more complaisant.
The most controversial stabilization, as well as the most significant in terms of its international implications, occurred in Germany after the creeping inflation of the war and immediate postwar period and then the hyperinflation of 1922–23. In stabilizing Germany, international politics loomed large. In 1923 France tried to detach the Rhineland from Germany by making overtures to German businessmen and drawing up a plan for a separate Rhineland currency. The French government hoped thus to use the might of finance to answer its security concerns. The German government responded to the French move by treating currency stabilization, for the first time, as an issue of pressing importance and by begging for British assistance. It approached the London capital market and set up a new central banking institution with a capital base in pounds sterling, at a time when the pound had not yet returned to a fixed gold parity.
The superior financial power of the United States and the evident German need for external capital inflows destroyed the prospect of a Mark stabilized on a sterling base. The eventual scheme adopted on the recommendation of the Dawes Committee of 1924 envisaged a return to a gold currency and assistance in the form of an international dollar loan.21 This marked the beginning of an era of dollar hegemony. The American financier Paul Warburg, one of the founders of the Federal Reserve System, wrote to Owen Young, a member of the Dawes Committee: “The opportunity that the present emergency in Europe offers is unique, and I don’t believe that it will ever again be within as easy a grasp of the United States as it is today. It is the question of whether the Dollar shall permanently retain a dominant position, or whether we are willing to surrender financial mastery to the Pound Sterling for good and all.”22
The German stabilization with a new currency, the Reichsmark, on a dollar base, together with an announcement that South Africa, then the world’s major gold producer, would return to a gold standard by May 1925, forced Britain’s hand. On 25 May 1925 the pound returned to the prewar parity against the dollar of $4.86. Sterling stabilization meant the effective end of a period of financial chaos in which only a few currencies—those of Mexico and the United States—had been linked to gold. The Belgian franc returned to gold in October 1925, the Danish crown and the Italian lira in 1926. In August 1926 the French franc was pegged, and a legal stabilization followed in June 1928. By the end of 1928 the gold-exchange standard had spread to thirty-one countries.
The gold-exchange standard represented the best way of providing such guarantees of stability as might allow large international capital flows to occur. Debates about whether the gold standard was an optimal system of management of international payments often miss the point that investors demanded such a system, with its stability and its constraints on the operation of sovereign monetary and financial policies. Its attractions lay precisely in the limitations it imposed on policy options.
International Central Bank Cooperation
Norman and Strong engineered first a system of informal cooperation among central banks. Later that cooperation was criticized for obliging countries and central banks to act against their own national interest for the sake of a vaguely conceived good of the international order, which would be soon undermined by specific national problems. In particular, in 1927, at a private meeting at the house of the U.S. Treasury secretary in Long Island, the European central banks persuaded Strong that Europe desperately needed an interest-rate reduction. Strong’s acceptance of this argument was later criticized as a cause of the asset bubble of the Wall Street boom.
Such cooperation rested on the precarious footing of the personal sympathies of Norman and Strong. When Strong died, in October 1928, Norman was frightened of financial destabilization. He tried to reach out to Strong beyond the grave, and invented the Bank for International Settlements as a way of achieving this economic spiritualism.
The Bank for International Settlements (BIS) had two purposes, which its founders may have intended to be complementary, but which proved instead to be quite contradictory. On the one hand, the Bank was supposed to end the politicization of the reparations issue, which had plagued the international financial system of the 1920s, and to provide a neutral, “market” solution. On the other hand, the Bank would act as an instrument of central bank cooperation, making the international capital markets less volatile. In effect, as a central bankers’ central bank, it was intended as a sort of world central bank.
The fact that this was in practice a “reparations bank” ensured that France and Britain would be locked in conflict about its role and function. Norman, more than the British Foreign Office, regarded reparations as pernicious and saw the Bank as a valuable instrument in demonstrating the absurdity of the entire concept. France—and the Banque de France shared this sentiment—saw the Bank as a means to guarantee the continuation of German payments for French reconstruction until the date (1988) established in the Young Plan adopted by the international conferences in The Hague. This conflict poisoned Franco-British discussions of monetary policy, and more generally—and with very long-lasting effects—brought about an intellectual bankruptcy in discussions of monetary policy. For a long time, central banks and central bank cooperation were associated with the terrible failures of the depression era. This “lesson” of the depression influenced the design of the Bretton Woods order; and it was only in the 1980s and especially in the 1990s that an assertion of the value of independent central banks reappeared generally (Germany and the Bundesbank served as a model in this discussion).
The Young Plan for settling German reparations in 1929 replaced the previous mechanism for the transfer of reparation payments through an agent-general, who was responsible for converting the Marks paid by the German government into foreign exchange and for making a judgment as to whether the foreign-exchange market would allow such a large transaction. Instead, Germany was to pay her reparations Marks to a new institution, the Bank for International Settlements. The BIS replaced the transfer-protection mechanism of the Dawes Plan through its discretionary power to reinvest reparation payments in German securities, and thus to remove pressure from the exchange rate. The Bank also acted as the fiscal agent for the Dawes and Young loans, as well as other international loans (the 1930 International Loan of Austria).
The BIS, however, was intended as rather more than a reparations bank. Its founders saw it as a way of mending the international order: stabilizing money, and providing depoliticized solutions to economic problems. Sir Charles Addis, a member of the Organization Committee established at The Hague conference to design the new bank, wrote: “it was hoped by this plan to fulfill the dream of Genoa by the gradual development of the BIS into a cooperative society of Central Banks, the governors of which would regularly meet together in concert in order to exchange information, and to devise means for promoting economy in the use of gold and for preventing by a common policy undue fluctuations in its value.”23 Later the objectives of the Bank were described as collaboration to “evolve a common body of monetary doctrine,” to “smooth out the business cycle, and to contribute toward a greater equilibrium in the general level of economic activity.”24
Montagu Norman formulated a very ambitious program as a way of implementing these objectives. He saw the prime tasks of the bank as lying in the “centralization of international monetary relations. It would act to prevent excessive credit leading to “overproduction when prices are artificially maintained (rubber etc.).” (Norman was thinking of the abortive Stevenson scheme, which had made rubber exports dependent on the price, had briefly raised rubber prices, and then led to overplanting and a catastrophic price collapse during the depression years.) The Bank would thus attempt to restrict the excessive amount of short-term capital moving internationally. One common diagnosis of the ills of the 1920s contrasted the long-term nature of prewar international capital movements with the volatile short-term flows of the 1920s (a debate reminiscent of some analyses of the ills of the 1990s). “To attract short-term capital to long-term markets is another task which can only be accomplished by identifying the policies of the Central banks, by coordinating the movements of their discount rates, by increasing the control of each in its own market.”25
France agreed about some of these goals. The French expert Pierre Quesnay saw the desirability of centralizing the statistical work of the various central banks in order to know more about the problems raised by international capital flows. But French thinking went much further and proposed that the BIS adopt a new gold currency (“grammor”) as a unit of account. The idea, characteristically French, goes back intellectually to the proposals of Emperor Napoleon III for a world monetary standard at the 1867 International Monetary Conference. Stripped of the gold element, however, it also looks forward to Keynes’s discussion of an artificial international currency (“bancor”) in the negotiations preceding Bretton Woods. The result would be that the defense of a currency in the case of a speculative attack would not require the sales of another currency (and hence the likelihood of transmitting the attack elsewhere).26 There were enormous hopes. The BIS in fact was the last great attempt to establish international economic cooperation before the Second World War.
The Bank’s statutes stipulated its responsibility “to promote the cooperation of Central Banks and to provide additional facilities for international financial operations.” It began operations on 17 May 1930 with an initial capital of 500 million gold Swiss francs, subscribed by central banks or (in the case of Japan and United States) banking groups.27 (To give some idea of the contrast in size with more recent institutions: the capitalization of the BIS amounted to 0.107 percent of 1930 U.S. GNP; the capital of the IMF was 4.019 percent of U.S. GNP in 1945.)
Its constitution, however, represented a rather political sort of compromise. The Paris experts, the Hague conferences, and the Organization Committee left the BIS, in the words of one of its directors, “vague, obscure, badly arranged and sometimes inconsistent.”28 In the first place, its membership betrayed clearly its origins as a reparations bank: it excluded all of South and Central America, Africa, the British overseas dominions, and Asia, with the exception of Japan, which owed its inclusion to its status as a (very small) reparations creditor. In Europe, Spain was left out. The United States, however, was brought in, though the representation was inevitably unofficial since the Federal Reserve System was forbidden to participate (because of the risk of involving the United States officially in the reparations quagmire). As a consequence, the BIS held its dollar deposits at two leading private New York banks.
The Bank was not located in any major financial center. The choice of site initially lay among the small countries of Europe, Belgium, the Netherlands, and Switzerland, with France strongly advocating a Belgian location and Britain and Germany equally militantly opposed. In Switzerland, the eventual choice, Zurich was rejected because although it was a major financial center, it was too German; Geneva involved too much of an entanglement with the League; and thus the choice fell on Basle. Norman had actually urged an even more peculiar Swiss choice, Bern, which had “the advantage of being a diplomatic, university and scientific centre and less of a money-making atmosphere”: the intention of maintaining the clublike atmosphere of 1920s in a rarified air was clear.29 Basle also had the advantage in the railway age of being at the intersection of the major European routes, London–Hoek van Holland–Rome, Paris–Vienna, and Berlin–Madrid.
The staffing took place in accordance with the principle of national representation. The first president of the Bank, Gates McGarrah, was an American; but the general manager in charge of the actual operation of policy was an extremely talented young Banque de France official, Pierre Quesnay, entirely dedicated to French national interests. German protests (especially from the Reichsbank president, Hjalmar Schacht) that he had been the figure responsible for organizing a speculative attack on the Mark in the spring of 1929 were ignored. Quesnay in fact had a powerful claim to his new position. Owen Young, the architect of the new reparations plan, hailed the thirty-six-year-old economist as the principal author of both the Young Plan and the Bank.30 As a means of conciliating Germany, Quesnay’s deputy was a German, Ernst Hülse from the Berlin Reichsbank. He proved a blinkered and unimaginative bureaucrat, more intent on warding off invasions of his administrative turf than on rescuing the international financial system.31 The result would have been a complete deadlock or a descent into routine and trivial business had Quesnay not possessed rather more imagination and initiative than Hülse.
It was difficult after the deliberations of the Organization Committee to avoid the conclusion drawn by a later British director of the Bank, Sir Otto Niemeyer: “No one who started out to construct a Super Bank for world cooperative purposes could conceivably have hit on the constitution proposed for the BIS.”32 The capital of the Bank came from the participating banks of issue. When the BIS began operations, its resources were so limited that the banking policy soon ran into a dead end: within months, by August 1930, the BIS approached complete illiquidity at the very time that the signs of world deflation and depression had become quite obvious.33 In the first year of its activity, the BIS had 1.8 billion Swiss francs in deposits, of which 300 million were reinvested in Germany, 650 million were short-term deposits by the reparations creditors who had not yet transferred their annuities, and 800 million represented other central bank deposits. Its only business that coincided with Norman’s vision was a stabilization credit for the Spanish peseta of £3 million in April 1931, which was designed to allow Spain to return to the gold standard (in fact the world financial crisis intervened).
The Bank was not permitted to make medium- or long-term investments (outside Germany) of the kind that might have been needed in drawing up stabilization packages. One of its staff now came to the conclusion that “if things continue to take their present course, the Bank will be in a completely frozen position within a month and unable to meet its liabilities without borrowing.”34
The urgent need for middle-term credits arose out of the world depression, which immobilized many bank loans: this was an instance in which a well-capitalized de facto lender of last resort might have played a powerful role in freeing the world from the incubus of frozen debt and illiquid banks. A subcommittee of the BIS in the autumn of 1930 started an inquiry into how the Bank might make up the shortfall caused by the growing bank problems of central Europe:
The [commercial] banks … are no longer prepared to continue this custom [the central European tradition of making long-term credits to commercial and above all industrial borrowers], which, from the point of view of rigid banking principles might be called an abuse, as, owing to the post-war economic depression these credits have become frozen almost everywhere, with the result that the banks are no longer prepared to invest money in companies with which they have already invested large sums not to mention the further fact that this freezing of credits has transformed a considerable portion of the liquid funds of the banks into fixed investments.35
The subcommittee recommended that a sum equivalent to the BIS’s capital, in addition to some permanent deposits, be placed in medium-term bills bought from banks in order to thaw central European credit. A more ambitious variant of the scheme appeared in February 1931 from the Bank of England and became known as the Kindersley scheme (Sir Robert Kindersley was a director of the Bank of England and of the BIS). It aimed to overcome the failure of international bond markets, in which—because of the collapse in security prices—new issues had become practically impossible. Kindersley and Norman recommended the creation of an international corporation with a capital of £25–50 million, which might issue bonds up to three times its capital to “foreign governments, municipalities, mortgage banks, harbor boards, railways and public utility companies.” “At a period like the present, when the capitalist system is largely under the microscope and is being attacked from many sides, it is of the greatest importance that capitalists as a whole should thus make an effort to find a remedy for at least one important difficulty which faces the money markets of the world today.”36
In fact the scheme, which attracted German support—since there the danger of financial collapse became ever more acute—found the French hostile and suspicious that this was an attempt to chip away at France’s political advantage arising out of the strength of the French capital market. The governor of the Banque de France, Clément Moret, argued that BIS participation in the Kindersley scheme would be contrary to the Bank’s statutes. Moreover, it was French banks that were supposed to subscribe most of the bonds under the scheme, “without being given the means of controlling the use of the funds furnished.” The debtor countries had only themselves to blame for the current weakness of international capital markets: “If a number of borrowers at the present time do not possess all the desirable facilities for procuring the capital of which they are reasonably in need, this is mainly because in the course of previous years too large a number of them have not strictly kept the engagements which they had undertaken with respect to their creditors.” There could be no point in relying on guarantees given by a borrowing state, since “in practice the creditor is powerless before a defaulting State; he comes into conflict with the ‘sovereignty’ of his debtor, and the political evolution of the last few years seems to have strengthened the force of this conception. The security given has only a very relative value and generally no value at all (for example Mexico, Turkey …).”37 The perils of sovereign lending became obvious to all during the depression.
Moret’s view seems plausible in retrospect. The first defaults came in Latin America. Bolivia had let its currency slip against gold in October 1930, and in January 1931 defaulted on its debt. Peru followed in March, Chile in July, and Brazil and Colombia in October. There then came the central European defaults: exchange control in Austria, Hungary, and Germany in 1931, and defaults by Hungary, Yugoslavia, and Greece in 1932 and by Austria and Germany in 1933.38 There was growing skepticism about sovereign loans. By the mid-1930s the liberal Swedish economist Per Jacobsson, chief economist at the BIS, was writing: “Political influence in lending is, as a rule, very costly; when a government has to put its influence behind a loan, the likelihood is that there is something wrong with the security of the loan.”39 But Moret’s pessimistic analysis does not take into account the possibility that early action might have limited the extent of financial contagion.
It was not, however, merely French opposition that brought down the Norman-Kindersley scheme. The American financiers were not sympathetic to a large-scale rescue operation. BIS president McGarrah cabled to Morgan partners Thomas W. Lamont and Seymour Parker Gilbert that the proposal was impractical and that it would have been much better to organize an investment trust through private banking channels. The Morgan bankers agreed with this assessment.40 Thus the proposal disintegrated, and Governor Norman noted sadly: “The fact is that the BIS is already slipping to the bottom of a ditch and in that position seems likely to do no more than helpfully perform a number of routine and Central Banking operations.”41
A more modest, but in some ways more interesting, proposal made by the middle-term credit subcommittee under the chairmanship of the influential Belgian commercial (not central) banker Emile Francqui, for the rediscounting by the BIS of commercial paper up to £10 million in order to prepare the way for a semiprivate corporation to be built up by the speculative Swedish financier Ivar Kreuger, fared little better. The idea was that rescue efforts involving central banks and official institutions alone would be doomed to failure. It was essential to “bail in” (to use more modern terminology) the private sector. But Francqui’s initiative was not at all well received by the two hostile camps in BIS policymaking. On the one hand, the British and Germans at the BIS regarded the idea as inadequate and limited; on the other, Moret described it as “utopian,” since “an issue of bonds at the present moment would, to say the least of it, be difficult.”42
Governor Moret’s pessimism was not unjustified, since the BIS Board meeting at which he delivered the death blow to the Francqui as well as Kindersley plans took place one week after the collapse of the Vienna Creditanstalt. The central European credit crisis now set in: the Viennese panic brought down banks in Amsterdam and Warsaw. In June and July the scare spread to Germany, and from there immediately to Latvia, Turkey, and Egypt (and within a few months to England and the United States). Less than one year after he was supposed to devise a scheme that might rescue the central bankers and the central European banks, Ivar Kreuger killed himself in a Paris apartment.
The second half of the 1920s looked as if they justified the hopes in the gold-exchange standard. From 1924 to 1930 $9 billion (and possibly as much as $11 billion) flowed, 60 percent of this sum coming from the United States. The United Kingdom lent some $1.3 billion and France $1.34 billion over the same period.43
Most of the flows from Britain and—more significantly in quantitative terms—the United States took the form of long-term capital bonds (see Table 2.1).
The overall capital flows in the interwar period were considerably lower than those of the prewar period and do not really justify the frequent description as an orgy of overlending. This fact becomes apparent once we consider the direction of lending and the flows from industrial to developing countries. For 1911–1913, the average annual capital export of Britain, France, Germany, and the United States to the rest of the world was $1.4 billion. In 1924–1928 this flow of capital to the developing world dropped to $860 million, or in price-deflated terms $550 million. In other words, if Germany—as a major recipient of the capital flows of the 1920s—is removed from Table 2.1, the stream of international lending looks rather modest.44 And the reasons for German borrowing were highly peculiar.
The shape of international capital flows in the 1920s and 1930s, however, looks similar to the boom-bust episodes that were characteristic of the nineteenth century or the restored capital markets after the 1970s. A flow of capital to debtor countries was followed by a collapse of confidence and then by a period in which the direction of the capital flow was reversed. Capital in the second phase returned to the creditor countries, and debtor countries were forced into adjustment (see Figure 2.1). For both Britain and the United States, the peak year of capital outflow was 1927. After that the U.S. collapse was much more dramatic, and after 1931 capital long-term outflows practically ceased. Britain still exported capital, but mostly to the empire and dominions.
Source: United Nations, Department of Economic Affairs, International Capital Movements during the Inter-War Period (Lake Success, N.Y., 1949), p. 25.
Britain, however, was also a major short-term debtor (as it had probably been already before 1914). So, in the 1920s, were Germany (then the world’s biggest debtor) and the United States. The BIS estimated total world short-term indebtedness in 1930 at 70 billion Swiss francs or $13.5 billion, of which only $4.3 billion related to commercial transactions. Germany accounted for $3.9 billion, the United States for $2.7 billion, and Britain for $1.9 billion.45 Figures solely for banking liabilities, however, show a higher British than U.S. net liability in 1930 (see Table 2.2). Both Britain and the United States played a similar role: they converted short-term deposits into long-term lending.
The origins of the relatively high short-term indebtedness of Britain and the United States lay not so much in any domestic problems as in foreign inflows that followed political uncertainty in Europe and Latin America. It would be wrong to see in British indebtedness a sign of economic vulnerability or an early symptom of industrial decline. The deposits originated in the turbulent circumstances of the postwar European continent. In the social explosions and inflations, large amounts of capital fled—out of central Europe, but also out of France.
Table 2.2 Short-term banking liabilities, United States and Britain, 1927–1930 (billions of $U.S.)
Note: Liabilities = total short-term funds due to foreigners on banking account.
Source: League of Nations, Balances of Payments, 1930 (Geneva, 1932), pp. 165, 181.
As an example, the McKenna Committee in 1924, which set out to examine the extent of German capital abroad, produced the figure, almost certainly too low, of 6.75 billion Gold Marks (GM) ($1.6 billion). Germans bought foreign exchange, while foreigners in turn used the Marks they received in order to buy nominal assets, which frequently depreciated rapidly as a consequence of inflation. Foreigners’ deposits in the Berlin great banks, which were estimated at 31.3 billion Marks (1.8 million GM, or $429 million), were worth only 140 million GM by the end of 1922 and 30 million GM in 1923.46
Short-term inflows to Britain continued during the great credit boom of the second half of the 1920s. After the great crisis of 1931, however, the general direction of the flows shifted; a massive wave of capital flight—estimated by Charles Feinstein and Catherine Watson at $3.5 billion—went to the United States and Britain.47 At first the motivation was fear of economic crisis and renewed currency instability; but as economic crisis had its poisonous and corrosive effects on political stability, there came an increasing political fear, of the likelihood of European war.
The experience of central Europe in 1931 demonstrated to what an extent financial shocks could aggravate the business depression, and how much the behavior of money mattered in the course of depression. In central Europe, banking crises brought more general economic and also political collapses. Two sets of problems came together: the instability of banking and the political impossibility of balancing budgets.
The League of Nations Economic Committee described the “new phase” of the world depression that began after the spring of 1931. “The crisis thereupon took on a more specifically financial character. The disturbance, however, extends to all spheres of activity. The operation of the world’s economic organization rests on confidence; as soon as this disappears it undergoes profound disturbances and the evil spreads rapidly.”48
The peculiarity of central European banks was their close relationship to industry. Traditionally the Universalbanken had taken short-term deposits but given long-term loans (so-called Kontokorrentkredit) to business; they also held shares while waiting for a favorable moment for flotation on the Bourse. This style of banking predominated in the Habsburg empire and its successor states, in Italy after the 1890s, and also in Germany and Belgium (where the oldest universal bank, the Société Générale, had begun operating in the 1830s). Elsewhere, in Scandinavia or the Netherlands or Switzerland, banks also held substantial long-term industrial assets, but usually through intermediaries, subsidiaries, or trust companies.
The central European pattern was in the 1920s and even later often considered to hold advantages for long-term industrial development, in that banks were much better supplied with information not only about companies but also about business conditions in general than investors operating in a system built around a stock exchange. The better information allowed a more rational basis for investment decisions, and this in the long run stimulated and promoted economic growth.
However, the financial situation of all these countries with universal banking practices was severely disturbed by the postwar hyperinflations. In the case of the former Habsburg empire, territorial changes and the splitting up of the big Viennese banks added to the dislocation. When banks reopened after currency stabilizations, their balance sheets were usually reduced to between a third and a fifteenth of prewar levels. This initial opening was usually followed by a period in which deposits and credits grew quickly: a catch-up phase. Above all, the banks expanded on the basis of foreign loans, usually short-term, which they converted in the quite traditional manner into long-term credits or equity participations.
Companies financed their recoveries after stabilization through credits from banks. The disproportion here between developments in the capital-exporting countries and in the capital-importing countries of central Europe is striking. Whereas from 1926 to 1930 the indebtedness of business corporations as a proportion of owned capital remained steady or fell in the United States (from 65.9 to 60.9 percent) and Britain (from 58.0 to 50.3 percent), these ratios in central Europe increased dramatically. Hungary went from 69.3 percent to a dangerous 80.3 percent; Germany from 65.1 to 88.9 percent. In Italy, which had universal banks but no substantial capital inflows, the indebtedness ratio declined, from 98.6 percent to 83.2 percent, in the absence of imported capital. On the other hand, as the figures indicated, there was a substantial degree of borrowing that made the industrial and business credit structure highly vulnerable to debt deflation should prices fall.49
Such credit structures were vulnerable in two eventualities: if the external inflow of capital, with which expansion had been financed, were to cease; and if a price decline led to a collapse of the value of the banks’ equity investments and endangered their loans by reducing the value of the collateral.
The worldwide application of the doctrine of independent central banking as expounded by the Bank of England and the Federal Reserve Bank of New York increased the vulnerability of central European credit structures. In order to demonstrate the “independence” that constituted the prerequisite for the maintenance of a continuing inflow of capital, central banks had to show that they were restraining monetary developments. Yet this was difficult in economies in which commercial banks had so much leeway in granting credits. Reserve requirements at the central bank were rare: indeed in Hungary it was considered unusual and dishonorable for banks to keep deposits at the central bank or to discount bills there. In these cases the central banks could generate international confidence and approval only by staging little crises. The peculiar agony of central Europe followed from the combination of strong national banking traditions with new Anglo-Saxon principles of central banking.
The banking systems became even more vulnerable because of the intensity of competition in banking in the 1920s, and because of the effects of a merger wave. In Germany, the most significant merger was that of the Deutsche Bank with the Disconto-Gesellschaft in 1929.
In the spring and summer of 1931 three different central European crises converged. The Austrian crisis started as a banking crisis, which then became a foreign-exchange and fiscal crisis. In Hungary a fiscal crisis set off a foreign-exchange panic and then a banking crisis. In Germany a fiscal and banking crisis coincided and set off the foreign-exchange crisis. These were not “twin” but “triplet” crises.
Austria suffered doubly in the 1920s: it was overbanked because of the legacy of Vienna as the financial capital of the Habsburg empire, and because of the large number of speculative institutions set up in the inflation. After the split-up of the empire, the Vienna banks were often left with the poorest assets in the successor states, while newly created national banks there took the investment plums. The stabilization of the Austrian currency was followed by bank collapses and mergers. In 1926 the Vienna Creditanstalt took over the Austrian business of the Anglo-Austrian Bank, and the Bodenkreditanstalt took over the Allgemeine Verkehrsbank und Unionsbank. In 1929 the Bodenkreditanstalt, which could survive only by rediscounting its bill portfolio at the national bank, was merged, at the instance of the government, with the Creditanstalt. This merger set up an Austrian colossus: according to a recent estimate, 60 percent of Austrian industry was dependent on the Creditanstalt.50
A crisis could have arisen at any time after the merger with the Bodenkreditanstalt. The moment came by chance, when one of the Creditanstalt’s directors, Zoltan Hajdu, suddenly, indeed inexplicably, in 1931 indicated that he would not sign the balance sheet “until the usual method of drawing it up was changed.” In the spring of 1931 the publication of accounts was delayed while Hajdu insisted that there be a comprehensive examination and revaluation of the bank’s assets. In the course of this reexamination it became clear that a revaluation would mean insolvency.51
Thus the first difficulties in central Europe appeared as early as 1929, in reaction to the lower capital inflows of 1928 and falling prices. In Italy the first bank problems began in the summer of 1930, and the affected banks depended on advances for the central bank (the Bank of Italy) to survive. From December 1930, deposits in the largest three Italian banks fell, and the movement became a near panic in April 1931.
Austria was remarkably and surprisingly calm until the spring of 1931. There had been a few small failures in 1929, but the general consensus was that these had not been enough to purge Austrian credit. “In spite of recent failures,” The Banker noted, “there are still too many banks in Vienna, expectations of whose development as an international financial centre have failed to materialize.”52 The announcement of a customs union between Germany and Austria, followed by the French protest against this démarche, increased nervousness; but there do not appear to have been any significant withdrawals of foreign short-term credits. The shock came suddenly: the Creditanstalt announced a delay in the publication of its accounts, and then, in the night of 11–12 May, revealed losses of 140 million schillings, which it attributed to the costly aftermath of the absorption of the Bodenkreditanstalt. Before 11 May, most foreign creditors had not realized what was occurring: but thereafter the affair became highly political. Depositors lost confidence in the Creditanstalt. By the end of May the bank had lost 200 million schillings in deposits. But only a quarter of this sum was deposited with other banks; the rest moved out over the exchange.53 As a run on the schilling started, the Austrian exchange was threatened, and Austria appealed for international help.
Governor Norman staged a rescue operation that was specifically intended to prevent the French from using the Austrian position for foreign-policy advantages. But Norman was also aware from the first of the dangerous international financial repercussions of the Creditanstalt case (what we now call contagion). “Nor must we forget,” he cabled to the New York Federal Reserve, “that a monetary breakdown in Austria might quickly produce a similar result in several other countries.”54 It took two weeks of tense negotiations to provide what was in the end a token amount, and which did nothing to restore confidence in Austria or in any other country.
The problem lay in the French response to Norman. The governor of the Banque de France, Clément Moret, knew, on the basis of information supplied from Basle by Pierre Quesnay, in this matter quite assiduous in the pursuit of France’s national interest, that the London market was too weak to help Austria. The London Rothschilds could not afford to support the Creditanstalt: “It can thus be foreseen that the Austrian government will sooner or later be obliged to sell its shares to a private group. In this respect it appears that the London Rothschild house will not be capable of acting. M. Quesnay announces the possibility that this offers to interested French banks.”55
Norman’s initiative resulted in two central bank loans organized by the BIS, though in July Moret tried to block the second Austrian loan because he could now argue that the international capital market had been so destroyed that it would be impossible to float a bond issue to pay off the loan.56 But these loans were a classic case of too little, too late: the initial 100 million schilling ($14 million) loan did not even correspond to the first, grotesquely minimal, estimate of the Creditanstalt’s losses. Charles Kindleberger’s verdict is on the mark: “The niggardliness and the delay proved disastrous.”57
The Creditanstalt was also an enormous domestic Austrian problem. The government rescued the bank through the purchase of bills by the national bank. On 16 June, as panic conditions spread through central Europe, the Creditanstalt’s creditors reached an agreement to keep the bank going, as otherwise the whole Austrian economy would collapse. At this stage the losses were calculated at around 500 million schillings, and the national bank had 690 million schillings’ worth of Creditanstalt paper. But even these figures underestimated the extent of the losses, which became apparent only in the course of an audit: at the end of 1931 the losses were reckoned to be 923 million schillings, or 725 million schillings more than the nominal capital and reserves after the government-inspired May 1931 reorganization of the bank. The assets included frozen loans to Austrian and central European industry. For 31 May 1931, the audited accounts produced by the London firm of Binder Hamlyn showed a total engagement of the bank of 1.339 billion schillings; the largest elements consisted of loans to the textile industry (11.9 percent of the total), petroleum (10.3 percent), automobiles (9.6 percent), state and local government (4.2 percent), and above all other banks (16.7 percent).58
By the spring of 1932 it became apparent that the Creditanstalt was by no means the only problem in Austrian banking, and that the other two large banks—the Niederösterreichische Escompte-Gesellschaft (NEG) and the Wiener Bankverein—had their own and equally grave problems. The representative of Austria’s foreign creditors complained about the “culpable frivolity” of the management of the NEG;59 and in the end he pushed through a merger of the Creditanstalt and the two other banks on 25 May 1934, with the frozen (and mostly worthless) industrial assets of the NEG separated out into a state-owned holding company, Österreichische Industriekredit AG.
The Creditanstalt catastrophe turned into a national emergency because of the immediate effect of the bank losses on the national budget. From the beginning, there had been a close association between the views of foreign bankers on Austria and the state of public finance. The cost of the Creditanstalt rescue alone amounted to 9 percent of Austria’s annual GNP. When this is added to the cost of other bank rescues, including that of the vulnerable system of publicly owned postal banks, the cost was comparable to that of bank bailouts in 1990s crises. (For Mexico in the 1990s, for instance, the rescue of the banking system cost 14.5 percent of GDP.)60
The Austrian budget had already been swollen by the rapid expansion of state initiatives in the later 1920s. Partly the new expenditure simply followed from the emergency program of the stabilization: the burden of pensions for the dismissed officials mounted continually. But the budget deficits were also driven by increased spending on welfare and on capital investment projects. Funds for these outlays seemed to be available because of the high tax levels established at the time of the stabilization in 1922–23. From 1923 to 1929 federal spending rose by 88 percent (from 6.3 to 9.1 percent of GNP), and local government spending by 169 percent (from 3.1 to 6.5 percent of GNP).61 Since the political dynamic to increase spending could not easily be reversed, short of massive external pressure on Austria, falling production and prices during the depression led directly to large deficits. The severity of the Austrian slump was also, however, a product of the high levels of government spending: the combination of high benefit levels and a high degree of unionization kept wages high, despite poor productivity growth in the late 1920s and falling productivity in the depression.62 Thus the development of large deficits during the depression functioned as an alarm signal for the whole of the Austrian economy (see Table 2.3).
In early May 1931, before the Creditanstalt difficulties were publicized, opinionmakers such as the Morgans in New York were alarmed by predictions of a large budget deficit for 1931.63 In France, even socialist newspapers were outraged by the size of the “hypertrophied” Austrian state.64 The bank rescue obviously widened the deficit, and threatened the capacity of Austria to service the League of Nations 1923 loan. In consequence, in January 1932 another Dutchman, Rost van Tonningen, was appointed by the League Financial Committee to supervise the Austrian budget. He devised schemes to raise taxes, cut civil service and railway staffs, salaries and wages, and pensions; but despite the imposed austerity on 1 July 1932 Austria halted the service of the League loan, and the foreign creditors were obliged to accept a new agreement (the Lausanne Protocol), which included a conversion loan secured on the gross receipts of Austrian customs and the tobacco monopoly, as well as a new domestic austerity program that narrowly passed the Austrian parliament, the Bundesrat, with a majority of one vote.
Table 2.3 Austrian budgets, 1928–1934 (millions of schillings)
Source: League of Nations, Public Finance, 1935: Austria (Geneva, 1935).
Hungary had been weakened by the collapse of raw material prices, since foodstuffs and raw materials constituted 59.9 percent of Hungary’s exports in 1929. Wheat alone represented 10.8 percent of the value of Hungary’s exports, and animals another 10.5 percent.65 Instead of bringing higher export returns, the superabundant harvest of 1929 simply depressed prices and contributed to the growth in stocks, with the result that in 1930, with a much poorer harvest, the price still fell. The average wheat price in 1927 had been 31.87 pengö/quintal, in 1930 19.11, and in 1931 12.78.66 Hungary had also borrowed extensively, and by the early 1930s the external debt service amounted to an impossibly high 16 percent of national income, or 48 percent of the value of exports.67
In the story of the Hungarian collapse, the mix of banking failure, crisis in public finance, external (exchange) problems, and foreign policy entanglement was very similar to Austria’s; but these problems were resolved in a rather different way.
Agricultural support by the government formed part of the problem, since it was the cash situation of the Hungarian treasury that precipitated the crisis in the summer of 1931. As early as 1930, Hungarian money started to move abroad nervously: the total losses for the year amounted to 70–80 million pengös. “Though all the facts are known only to a very few individuals, there are signs that public confidence is not to be relied on,” the National Bank governor Sándor Popovics confided to a Bank of England official in February 1931.68
As in Austria, uncertainty over the public budget was a critical element in the collapse of confidence. Here, however, the fiscal problem played a key part at the beginning of the crisis. The National Bank described the Hungarian difficulties in the following way:
the fact that such a situation could remain undisclosed until so late a stage argued a lack of financial control and administrative organization which could not be denied. Individual departments had got out of hand; they had misled the Treasury and incurred liabilities on their own authority. The Government as a whole had also been too liberal, for example in affording various kinds of relief. But the chief source of trouble was the failure of revenue on account of the agricultural depression.69
One major problem lay in the extent of the losses incurred by the government in the operation of the boletta, a wheat price-support scheme that suffered as the wheat price plummeted, and of the futura, a buffer stock of wheat.70
The difficulties in public authority budgets were, however, less significant than the problems of publicly owned enterprises. An extensive nationalization, coupled with the attempt to build up strategic industries and reduce dependence on imports, transferred the problems of industry in the world depression into difficulties in the state budget, and by this route straight to the attention of the international financial community.
From 11 May 1931, the date of the open Creditanstalt crisis, Hungary’s National Bank lost foreign exchange at a rate of 150,000 pengös a day. The National Bank attributed the losses to a combination of American withdrawals and Hungarian capital flight: Hungarians were buying Hungarian shares domiciled abroad after the failure of the Austrian bank and security house Auspitz Leben & Co. in the wake of the Creditanstalt crisis.71 In order to finance these purchases, Hungarians withdrew deposits from banks, and then the general panic set in. Current accounts in the twelve largest Budapest banks and in the postal savings banks fell from 757.6 million pengös in May to 637.1 million pengös in October; and as banks withdrew credits to meet these demands, the number of bankruptcies jumped. This was the Hungarian version of debt deflation.
The National Bank kept the banks alive by generous rediscounting before a public collapse of the credit institutes, unlike in the Austrian case. This generosity allowed the outflow of funds across the exchange to swell, so that the strain fell on the external exchange rather than directly on the Hungarian banks. Initially, this strain was managed with the help of the BIS: on 6 June the BIS gave a $5 million credit, and on 11 June a further £1 million (approximately $5 million) followed; a third $5 million loan went out on 22 June. But this did not look like solving very much, since the BIS, and the foreign creditors, could calculate that over the next three months, even without a banking panic, £5 million in credits were due (of which £3 million was owed by the government).
A substantial part of the problem lay in foreign indebtedness. The total obligations of the Hungarian banks amounted to 483.21 million pengös, of which 232.26 (48 percent) was owed to U.S. creditors and 104.95 (22 percent) to British creditors.72 The foreign creditors negotiated a standstill agreement that became a model for subsequent similar accords with Austria and Germany. It was a general framework, and it took longer to elaborate the particular details, with the British banks reaching a settlement on 14 March 1932 and the U.S. banks on 21 June. At the outset of the crisis, the Hungarian government complained that “the United States had not yet gone through the school which a leading nation had to pass through before it could count on the confidence of all its customers.”73 The United States, in short, was not acting as a beneficent hegemon.
As in Austria, the foreign credit situation appeared in Hungary to require the imposition of austerity. In the autumn of 1931 Hungary drew up an eleven-point program for the League of Nations, which included the predictable list: civil service pay and pension cuts, tax increases, the ending of state subsidies, and in particular the abolition of the boletta and futura. Civil servants indeed had their pay docked, but the agricultural support schemes continued. The cutbacks are clearly discernible in the public administration budgets of 1932–33; but it is also clear that public undertakings were affected less severely. The boletta operated until July 1934, by which time wheat prices had substantially recovered. Under agrarian pressure, the government in 1932 and 1933 introduced extensive legislation to make farm bankruptcy impossible.
The most dramatic and extensive central European banking crisis occurred in Germany, where on 13 July 1931 the Darmstädter und Nationalbank (usually called the Danat Bank) shut its doors, the government declared a bank holiday, and the entire financial system required reconstruction with costly government assistance. The origins of the crisis were very similar to those in Austria and Hungary: the vulnerability of banks because of recent hyperinflation, the flight of domestic capital, the reaction to political uncertainty (both domestic and international), pressure by farmers for more effective protection against bankruptcy, and the triggering effects of relatively small cash deficits in the public accounts.
German capital flight predated the world depression. The origins lay in the war, and above all in the immediate postwar period, when a large number of German corporations founded and used subsidiaries in Switzerland and the Netherlands to move capital out of Germany. Another wave came in the late 1920s, partly in order to escape taxation but also in response to continued insecurity about German developments. Before the onset of the depression and the disintegration of the German polity, from 1927 to 1930, the capital accounts reveal an outflow of 3.9 billion Reichsmarks (RM) in short-term capital and 4.9 billion RM in long-term capital, while the short-term liabilities of the United States rose.
In 1929 a large number of small German banks failed: Richard Harte, Berlin; Julius Cunow & Co., Berlin; Fritz Kienstedt, Lübeck; Kieler Bank and Bankhaus Horst Fritzsche, Dresden; Paul-Schlesinger-Trier & Co., Frankfurt. Most significantly in August there was a major collapse of an insurance firm, the Frankfurter Allgemeine Versicherungs AG, followed by runs on municipal savings banks in Berlin and Frankfurt. A foreign periodical commented in April 1930 on the aftermath of these little crises: “The crisis which caused the failure of innumerable commercial firms and a large number of small banks seems to have passed its climax, and it is believed that those banking and commercial firms which have survived the crisis may be regarded as safe and sound.”74 Unfortunately this prediction did not come true.
From May 1930, German banks lost deposits, and by June 1931 the money supply had in consequence fallen by 17 percent. The banks had substantial foreign short-term liabilities. In July 1931, after a substantial part of the foreign loans had already been called, the German banks still owed 5.9 billion RM, and total German short-term indebtedness amounted to 13.1 billion RM. The total foreign indebtedness one year earlier is estimated at 15.5–16.5 billion RM. Foreigners had treated Germany with greater suspicion since the government crises of December 1929, when the finance minister was forced to resign; March 1930, when the socialist-liberal Great Coalition collapsed; and September 1930, when the Nazi party achieved an unexpectedly high vote in the Reichstag election.
On the other hand, the problem for Germany lay less in the absolute quantity of foreign investment than in the double danger posed by the vulnerability of the German banks and the weakness of public finances. The banking system had expanded rapidly after the currency stabilization.
After the Creditanstalt crisis, German banks began to lose funds more rapidly. In May, the Berlin great banks lost 337 million RM (or 2.6 percent of their deposits); and in June the bank with the weakest reputation, the Danat, lost 40.9 percent, while the more solid Deutsche Bank lost 8.2 percent. Most of the initial withdrawals were not foreign, but were made by Germans; and it was also Germans who moved money across the exchange into the Netherlands and Switzerland. By the middle of June, however, the foreign banks had become highly concerned. On 23 June the Bankers Trust cut its credit line to the Deutsche Bank. On 3 July the governor of the New York Federal Reserve Bank called in leading New York bankers, representatives of J. P. Morgan, Lee Higginson, Chase, National City Bank, Blair & Co., New York Trust Co., and Central Hanover Bank and Trust Co. After some persuasion they agreed to “at least maintain their present position and in some cases indicated that they would even reopen unused credit lines.”75 But on 6 July the Guaranty Company of New York started to withdraw from Germany. It was, nevertheless, rather late in the day: a great deal of capital had moved before the Americans started to feel nervous.
Because the withdrawals and the capital flight took place across the exchange, the Reichsbank lost reserves. On 30 May the gold reserve stood at 2.39 billion RM; by 23 June it was only 1.421 billion RM. This was close to the minimum reserve of 40 percent of note issue laid down by the Reichsbank law, and the president of the Reichsbank, Hans Luther, appealed for foreign help, and in particular for central bank credits. The BIS arranged a $100 million credit from the Bank of England, the Banque de France, and the Federal Reserve Bank of New York. On 20 June U.S. president Herbert Hoover announced a one-year “holiday” on reparations and inter-Allied debt payments.
But the price for this help was an insistence that the Reichsbank restrict its discounts in order to make capital flight out of Germany more difficult. This is what Luther did on 22 June, but the measure completely failed to stop the movement out. Partly politics were to blame: the French refused to agree to the Hoover moratorium until early July, and thus international tension grew. But Governor Harrison of the FRBNY, as well as Norman, blamed Luther for not taking more effective steps to stop capital flight, which he believed to be at the heart of the German problem. Harrison cabled to Norman: “I felt that the chief difficulty was a flight from the Reichsmark by German nationals and that the Reichsbank should resort to much more drastic credit control than apparently was the case.” And again: “Rationing of credit is of course a drastic and disagreeable procedure but it has been applied effectively in Germany in the past without proving to be fatal. On the contrary, in each instance it has been most helpful in repatriating German capital and in checking further outflows of funds and I cannot see why it might not be equally effective at this time if applied with equal force.”76
The Germans were vulnerable because of the character of their financial structure. The shock to German confidence lay partly in the vulnerability of German banks: in their overcommitment to specific firms such as Nordwolle (the immediate apparent cause of the Danat’s difficulties) and the brewery Schultheiss-Patzenhofer. But there existed also a fear of public bankruptcy. This is where the political element—which played a major role in the German crisis—and the financial panic intersected.
The German Finance Ministry faced a cash problem: despite round after round of tax increases, revenues consistently fell below anticipated levels; and despite repeated economy drives, axing of capital projects, civil service pay cuts, and reductions in unemployment support, expenditures remained too high. In consequence, the German government needed to borrow money from the banks in order to make its regular payments. The reparations sums due on 15 June and 15 July looked like particularly insuperable burdens. But it became more difficult to borrow from the banks as the Reichsbank limited its discounts of bank paper; and banks, if they were to lend to the government, needed to reduce credits elsewhere. This in turn hit industrial and commercial borrowers. The large payments for mid-July hung over the increasingly frenetic debates of early July and over the desperate bid of Luther to raise another central bank credit.
On 4 July there were signs of an imminent collapse of the credit of some of the large cities of the Rhineland and Westphalia. The Danat had been particularly involved in municipal lending. On 5 July the Basle National-zeitung announced that one of the leading German banks was “in difficulties,” and a discussion of the Danat’s losses as a result of its involvement with Nordwolle began. After a weekend of round-the-clock talks, in which the Dresdner Bank revealed that it too was close to failure, on Monday 13 July all German banks were closed by government decree.
After this bank holiday, the state needed to intervene in order to reconstruct the German banking system. The Reichsbank and the banks founded an Acceptance and Guarantee Bank in July to provide an additional signature to make bank bills eligible for Reichsbank discounting: it was a tacit promise that the Reichsbank would introduce a more generous discount policy. This help allowed the banks to recommence operations on 5 August. But a restructuring of the banking system was still needed: enormous losses were written off with state participations in the new capitalization of banks. The Danat was merged with the Dresdner, and by 1932 91 percent of the Dresdner’s capital, 70 percent of the Commerzbank’s, and 35 percent of that of the Deutsche Bank was in public ownership. Other public institutions were formed to take over and write off bad assets and manage long-term industrial participations. Solving the problem of the bad banks was thus a general problem for policy in the capital-importing countries.
The factors that had led to the transmission of crisis in central Europe—namely a high external debt, falling export prices, government fiscal difficulties, consequent fears about debt service, panics, external drains, and internal banking crises—also produced a decade of misery in South America. But the course of subsequent events was not identical. The European countries on the whole stayed on the gold standard for as long as possible, and were “forced off” by financial and banking crises. On the other hand, Latin America and Australia quickly used depreciation as a response to commodity-price induced balance-of-payments difficulties.
Like central Europe, the region had been a major recipient of capital inflows in the 1920s, on terms very similar to those of the central European borrowers. One quarter of the new capital issues floated in New York for foreign borrowers went to Latin America, and $2 billion worth of bonds were issued in the New York bank market.77 Such loans were marketed very aggressively in the United States. By 1932 an estimated one and a half million individuals held foreign securities, and in 1937 the Securities and Exchange Commission estimated that 600,000 to 700,000 investors held defaulted bonds. After the financial crisis, as was the case with the central European bonds, the issuing and underwriting banks were accused of carelessness in the promotion of their bonds and of grossly underestimating the risks involved. The debate contributed to a widespread feeling that a fundamental reform of banking was needed.
Charles Mitchell, the chairman of National City Bank and one of those accused of misleading the public, informed the Senate Committee on Finance that “those bonds were bought by Tom, Dick and Harry … without reference to the solidity or the solvency of the bonds … but entirely on the faith of the house issuing them in New York.” Other bankers gave evidence of how American banks had used high-pressure tactics to sell loans to Latin American countries. There were twenty-nine bank representatives in Colombia. Thomas W. Lamont, a partner of one of the two banking houses that did not aggressively pursue such business, stated disapprovingly in a speech in 1927: “I have in mind the reports that I have recently heard of American bankers and firms competing on an almost violent scale for the purpose of obtaining loans in various foreign money markets overseas … That sort of competition tends to insecurity and unsound practice.”78
U.S. banks also engaged in substantial short-term lending to Latin America, both to governments and to corporations. There was too a substantial British engagement both in the long- and the short-term markets. Historically, Britain had been especially involved in the financing of railroads and municipal infrastructure, and this bias remained in the 1920s. In 1931 the British-domiciled bonds outstanding of dominion and colonial governments and municipalities amounted to $5.372 billion, and the bonds of foreign governments and municipalities to $1.643 billion.79 Vulnerable bank loans to South America played a decisive role in the critical weeks of the financial panic of 1931 in the United Kingdom.
For most of the 1920s, the loans appeared to be quite safe. There were few defaults, although the yields always lay above those of high-grade U.S. corporate bonds. The effective interest rate on long-term public-sector debt in 1927 was 6.5 percent for Germany, and 7.4 percent for Austria. The Latin American rates were exactly in this range, with the relatively prosperous and established Argentine republic paying 6.7 percent, Chile 6.9 percent, Colombia 7.0 percent, Peru 7.2 percent, and Brazil 7.5 percent.80 Subsequent, after-the-event analysis indicated that the loans concluded in the great waves of borrowing in the 1920s were actually progressively less and less sound (in that the subsequent default rates were higher); but this was of course not apparent at the time.81
As the debt built up, the debt/export ration deteriorated, so that by 1930 the external debt of Bolivia was 237 percent of exports, and for Chile 121 percent, levels that were repeated in the post-1945 era only in the great debt build-up of the later 1970s. In 1930, though, Peru at 76 percent and Argentina at 46 percent looked much more manageable.82
In 1929–30 two related shocks occurred: export prices of many commodities fell, and the inflow of fresh funds was dramatically curtailed. Both raised questions about the ability of borrowers to continue debt service. Wheat and some metals had been falling in price steadily since 1925, but after 1928 the decline became substantially steeper. Wool prices began to decline from the middle of 1929. Coffee reached a peak in the spring of 1929, and then the price fell by almost half by the end of the year.
As prices fell, export earnings were cut. From 1929 through 1931 Argentine export values fell 32 percent (the import side was reduced even more dramatically, as the end of the flow of foreign funds led to economic contraction). In Chile over the same period, with a heavy dependence on copper, export values fell 64 percent. Since a major source of government income was collected through levies on exports, the results for public finance became quickly clear. In the event, over the same period government revenue fell a relatively modest 8 percent in Argentina, but by 20 percent in Brazil and 36 percent in Chile.83
Another way of measuring the severity of the Latin debt problem is to examine the debt service expressed as a share of export earnings. For Argentina in 1927 this ratio had been 7.9 percent, and for Bolivia, 6.1 percent; for 1931 the equivalent ratios were 22.5 and 24.5 percent.84
The combination of the difficulty of maintaining debt service as new loans dried up, and the fall in export prices, quickly led to currency depreciation. This was a different story from that in Europe, which had been more central in the history of the gold standard, and where arguments about credibility played a greater role. Latin America had been marginal in the years of the “classical,” prewar, gold standard, in that there were relatively frequent suspensions of the exchange and departures from parity. A second consideration, which again diminished the weight of the credibility argument that was made so frequently and forcefully in the central European cases, was that there had been no recent experiences of extreme inflation. In consequence there was less of a need to cling to a nominal exchange anchor as a tool of stabilization policy. (To put this in terms of 1990s parallels: the situation of Latin America in the 1920s was closer to that of East Asia in the 1990s, where there was less need for an anchor, than to that of Latin American countries in the aftermath of inflations in the 1970s and 1980s.)
Currency depreciation was in many countries a fairly rapid response to falling commodity prices. Argentina and neighboring Uruguay suspended the gold standard in December 1929; at the same time the Brazilian exchange began to depreciate. In March 1930 the Bolivian exchange fell, and in September 1930 that of Venezuela. These movements were responses to trade problems, but they affected debt policy in that a consequence of depreciation was to make foreign debt service more expensive in terms of the national currency. As in central Europe, then, an attack on the exchange rate could be read as a threat to the stability of the debt structure.
Bank vulnerability constituted a final element of weakness in some of the more developed economies. The different experience of Argentine and Brazilian banking goes some way to explaining a different room for policy maneuver later in the 1930s. Argentina had experienced a large expansion of bank credit in the 1920s, and in the depression, undercapitalized banks were vulnerable to panic. In April 1931, weeks before the outbreak of the central European banking crisis, a major depositors’ run set in. The government response was different from that of the central European authorities: the quasi-official Banco de Nación (there was no central bank) was allowed to discount commercial paper (including, presumably, some bills that represented purely financial transactions) at the government-owned Casa de Conversión. But the extent of the Argentine banking panic can be judged from deposit statistics: from 1928 through 1931, deposits in commercial banks contracted by 25 percent. The government program had a major fiscal cost. Already in 1930, the government deficit was 40 percent of expenditure.85 Chile had a very similar experience (26 percent), while the Brazilian banking sector, with a contraction of only 5 percent, was barely affected.86
In 1930 some Latin American bonds fell sharply in price on the U.S. market, with both Brazil and Bolivia now priced below 50 percent of the nominal value. The rating agency Moody’s downrated Peru to Baa, and Bolivia, Brazil, and Venezuela to Ba.87 The first debt default came in a country with a very unfavorable debt/export ratio. In December 1930 Bolivia defaulted on the old government sinking fund, and in January 1931 on the general payment of interest. Peru, by now engulfed in civil war, defaulted in March 1931. Some countries (such as Chile), however, still even managed to raise long-term capital in 1930, as investors believed that the collapse of the copper price was only a temporary blip; but the country descended into political chaos, with a revolt in September. In March 1931 import tariffs were raised, but this measure was not sufficient to deal with the balance-of-payments deficit. In June 1931, following the Hoover moratorium on war debts, Chile suspended interest and service charges and imposed exchange controls. At the end of July the government resigned and was followed by a short-lived “Socialist Republic.” At this stage, a general regional panic began. Colombia defaulted first on municipal and departmental bonds, in 1931, and on central government debt in 1933. Cuba defaulted on $170 million in U.S. debts in August.
There was little political response to the chain of Latin American defaults. It was by now a story familiar from Europe. The U.S. government refused to intervene, and in fact saw its major future task as regulating U.S. banking to prevent the abuse of the North American capital market. A U.S. Foreign Bondholders’ Protective Council, created as a response to the calamity in 1933, was almost completely ineffective.88 During the 1930s some Latin American governments, notably Chile, bought back defaulted bonds at relatively low prices. The only major country that did not default was Argentina.
Debt defaults went hand in hand with devaluations (a consequence of the default was that there was less need to worry about the effect of devaluation in increasing the internal value of external debt). What was the room for policy maneuver? In an influential essay, Carlos Díaz-Alejandro tried to differentiate between larger countries or those with “relatively autonomous public sectors,” which could manipulate exchange rates, tariffs, and domestic credit. Small, mainly Central American states, on the other hand, such as Guatemala, Honduras, Nicaragua, and Panama, as well as Cuba, suffered because they had no autonomous exchange policy (they remained pegged to the U.S. dollar).89 El Salvador had a small depreciation against the dollar, but its economic recovery remained modest.
In many of the depreciating economies the measures led to a real depreciation (i.e., prices did not follow the devaluation) and provided the basis for an export-based recovery.90 In Argentina, Brazil, Chile, Colombia, Mexico, Peru, and Uruguay, real exchange rates were between 30 and 90 percent of their 1925–1929 levels. Wages, like prices, stayed low.91 The result was that substituting for imported goods became a much more attractive strategy. Colombia and Brazil already showed some measure of recovery in 1932; Peruvian cotton exports picked up in 1933. But the long-term basis for a really sustained export-based recovery was threatened by the increasing trade protectionism of the industrialized countries (see Chapter 3). Where there was no default, there was a much slower recovery, and Argentina stood out as by far the poorest performer of the major South American economies in the 1930s.
The central European banking crisis of 1931 had immediate repercussions throughout the world, particularly in the Middle East and Latin America, where an important part of the banking structure was in foreign ownership and thus very vulnerable to crises in Europe or North America. In Turkey in July 1931, there were runs on branches of the Deutsche Bank and the collapse of the Banque Turcque pour le Commerce et l’Industrie. In Egypt, the Cairo and Alexandria branches of the Deutsche Orientbank closed their offices. In Latvia, those banks with a German connection were hit: the Bank of Libau and the International Bank of Riga. What began in Rumania as a crisis of just the German-associated banks—particularly the Banca Generale a Tarii Romanesti—became a general panic: a run on the Banca de Credit Romana and the Banca Romaneascu. Before the end of the year, one of the leading banks, the Banca Marmarosch Blank & Co., collapsed.
In Mexico a leading bank, the Crédito Español de México, collapsed in July 1931, and the crash was followed by a run on the central bank as the confidence of foreign investors evaporated. There was the Argentine crash, and then the Chilean panic. In Italy no open panic occurred, but the intrinsic position of the banks was as weak as in central Europe. From December 1930, deposits in the three largest Italian banks fell, and in April 1931 depositors came near to panic. The state responded with a rescue operation, which transferred illiquid industrial assets to a new holding company, Istituto di Ricostruzione Industriale (IRI).92
These weaknesses contributed to the danger to the most international of the world’s financial centers, the City of London; and then, in the annus terribilis of international financial relations, London produced its own shock to the world system.
Conventional wisdom has it that some sort of sterling crisis, which would inevitably lead to the abandonment of the gold standard rate of £ = $4.86, followed more or less inevitably from the choice of an overvalued parity, the prewar parity, for prestige reasons in 1925. This is the view of John Kenneth Galbraith: “In 1925 began the long series of exchange crises which, like the lions in Trafalgar Square and the streetwalkers in Piccadilly, are now an established part of the British scene.”93 This view is repeated in most recent analysis: it was most effectively argued by Donald Moggridge, but appears also in the works of Sidney Pollard and Robert Boyce (who argue that $4.86 represented a victory of a “City” financial interest over a commercial and manufacturing one), and Diane Kunz (who takes the more harmless view that it was just a mistake).94
This view is difficult to sustain on the basis of calculations on the value of sterling based on either purchasing power parity or real exchange rates. Milton Friedman and Anna Schwartz give a purchasing-power parity for 1929 of $5.50, indicating a substantial undervaluation of the pound. British net exports bear out the same story, rising absolutely and in real value and relative to imports after the mid-1920s.95 Eichengreen has used an econometric model to show that the unfavorable development in the British balance of payments in 1931 (as Britain’s earnings from services and especially from shipping fell off during the world depression) cannot account for the severity of the exchange losses. Temin in consequence is obliged in his recent account to argue that the 1925 mistake was not the choice of exchange rate—the old story about the Norman mistake (the decision driven by the Bank of England’s idiosyncratic governor, Montagu Norman, to return to gold at the old parity)—but rather the choice of any parity: it was the commitment to a fixed exchange-rate regime that tied Britain’s hands. This is a much more radical objection to Norman’s policy, but it is one made by hardly anyone at the time.
Instead the explanation lies in the domino effects on the banking structure of the central European events. Britain and the United States were vulnerable because of their position in the international capital markets as major short-term debtors. Britain was the first to be hit by the panic.
In memoranda produced during the 1931 crisis, the Bank of England’s officials repeatedly referred to the high costs of staying on the gold standard. What did they mean? What were the high domestic costs that were being imposed as a consequence of maintaining the gold convertibility of sterling? We should refrain from adopting the anachronistic view that these costs lay in the forced reduction of public expenditure, the cuts in pay or the reduction in the dole. On the contrary, the Bank and its world were convinced that these were necessary—whatever the exchange parity—if a stable rate were to be maintained at all.
In fact the domestic costs lay in the vulnerability of the British financial system. This was already made apparent in one of the first Committee of Treasury (the critical decisionmaking body at the Bank of England during the crisis) meetings on the crisis. On 27 July 1931 the meeting was joined by a representative of the British Treasury, Sir Richard Hopkins, who had just set out a memorandum in which he formulated very clearly the British danger:
We cannot control that we are in the midst of an unexampled slump, nor the fact that Germany is bankrupt, that great assets of ours are frozen there, and that foreign nations are drawing their credits from there over our exchanges. Nor can we control the fact that foreign nations have immense sums of money in London and will try to get them away if distrust of the pound extends … the first thing at which foreigners look is the budgetary position.96
At the same time the British clearing bank (the British term for commercial banks) representatives met to hear the Bank’s view and to present their own demands.
On the twenty-seventh, Hopkins’ task was to present Chancellor of the Exchequer Philip Snowden’s view to the bankers: “If such credits [from France and the United States] are raised, and indeed in any present contingency, the Bank should be prepared to use its gold to the extent necessary and H. M. Government will be ready to increase the fiduciary issue to enable such gold to be released.” But he also added a warning about the penalties for failure: “If credits cannot be arranged and gold continued to be withdrawn, British banks must be entirely free to withdraw credits from Germany.”97 Yet the latter could not be a realistic option. The London conference had recently recommended the maintenance of foreign short-term credit to Germany on the insistence of the world financial community, and everyone realized that it would be impossible for banks to extricate themselves from Germany without bringing the whole credit structure crashing down.
Montagu Norman and his deputy governor, Sir Ernest Harvey, left the Committee of Treasury, went to a meeting of the clearing bankers, and reported back. The clearing banks, they said, “opposed the idea of any credit,” because a foreign central-bank operation to assist the Bank of England would allow the Bank to continue to make gold shipments that would be financed by the withdrawal of deposits in London banks. In short, the external drain would turn Britain into an Austria or a Germany.
It was this awareness, that the British problem lay fundamentally in the liquidity and solvency of some banks, that made the credits from France and the United States look like such a poor idea, and made the Bank unwilling to use either the Bank rate or its own reserves in the crisis. But since the problem lay in the banks and in the delicate area of confidence, it would have been entirely counterproductive for the Bank to make public this argument. Prime Minister Ramsay MacDonald also always spoke about a “flight from the pound” and a “financial panic” in the same breath: we know how the former was reflected in the exchange losses at the Bank of England, but the latter has been much more obscure.
English commercial banks had traditionally avoided long-term commitments to industry, and had indeed been taken to task for their neglect by the Macmillan committee (whose report on the weakness of the British system, published on 13 July, itself played a role in the British crisis). In fact, however, a deflationary environment, such as existed in Britain after the collapse of the postwar boom in 1920, can immobilize even short-term credits. During the 1920s British banks became ever more closely involved with the fortunes of Britain’s crisis-bound staple industries, especially Lancashire textiles. There are similarities here to the continental disasters of the Austrian Bodenkreditanstalt and the German Danat, in both of which textile lending played a major role. The Bank of England, in its role as financial policeman, therefore eventually became involved in the reconstruction of Lancashire and in regional policy through the Bankers’ Industrial Development Corporation.98
In addition to problems relating to domestic industrial commitments came in 1931 the continental situation. Foreign exposure was the Achilles’ heel of the London City. The position in this regard for the large joint-stock clearing banks was much safer than that of the private bankers, Schroeders, Lazards, or Kleinworts, who had committed themselves heavily to central Europe. Since the first outbreak of the continental difficulties, rumors had swept the marketplace about the position of British banks. On 18 May 1931, Governor Harrison had thought them sufficiently grave to cable to Norman: “For your information and such comment as you care to make, persistent rumors have today run the gamut here regarding Barclays Bank in particular and also Schroeders and lastly Rothschilds.”99 In view of events in central Europe, Kleinworts appealed to the Bank of England for a guarantee of its overdraft from the Westminster Bank (one of the large commercial—in British parlance, clearing—banks), but unsuccessfully.100 In addition to investments, the banks were threatened because of large forward positions from German, Austrian, and Hungarian persons and institutions.101 There was some awareness of the fragility of the merchant banks even before the crisis. A rating system in the autumn of 1930 evaluated the clearing banks and the large American banks as AAA, but rated Higginson & Co. (heavily involved in central Europe) and J. H. Schroeder as A, and Kleinworts as BI.102
The merchant banks tried to refinance themselves with the clearers. On 15 July, for instance, the Anglo-French Banking Corporation informed the Midland Bank that it was anticipating £700,000 in withdrawals and asked to borrow that sum. There was another mechanism that touched the clearing banks: the indirect effects of the central European crisis. The Banca Commerciale Italiana had £5.5 million in credits from the Midland, some of which reflected onward lending to Germany.103
At the Bank of England’s Currency Committee, established immediately in the wake of the devaluation, one banker argued that £100 million in deposits had been lost overall by the English system since the beginning of the year. Another banker said very candidly: “The Banks have great difficulty in making both ends meet. Their losses have been terrific.”104 For the nine clearing banks, the aggregate monthly average of deposits fell from £1.836 billion in January 1931 to £1.688 billion in October. The banks were also affected differently: in the course of 1931, Barclays lost 3.9 percent of deposits and current accounts, the Westminster 6.6 percent, Lloyds 8.5 percent, the National Provincial 10.4 percent, and the Midland (the largest English bank) 9.8 percent.
The Economist wrote that though the European standstill did not affect the joint stock banks, “several Accepting houses and especially some of the newer international banking houses founded in London during the past decade were seriously implicated, and the latter also had to face a sudden and wholesale withdrawal of their foreign deposits, which formed the bulk of their resources. With many of their assets immobilized, they had to meet the withdrawals by sales of securities at a time when the gilt-edged [government bills] market was abnormally depressed.”105
The most energetic and persistent pressure on the Bank to devalue thus came from bankers who feared for their own position. Most striking is the position of Sir Robert Kindersley, one of the most active and vigorous directors of the Bank, and chairman of the threatened bank Lazard Bros. & Co. The bankers also turned to the politicians. On 16 September 1931, Sir William Goode, who in the 1920s had served as informal adviser to the National Bank of Hungary, wrote to Prime Minister Ramsay MacDonald, stating that the gold standard could not be kept unless long-term loans “for the other countries” of Europe started to flow again and thus reduced the strain on British banks.106
It was not just central Europe that presented a threat to the English banks. The final blow to sterling in the judgment of foreign markets came with the announcement of British difficulties in South America. The Echo de Paris reported that “the news that the Brazilian coupons would not be paid on 1 October increased the disarray. England is the largest creditor of Latin America.”107
Concerns about the stability of English banks explain the otherwise mysterious failure to support sterling through intervention in the Paris market on 5 August: by letting the exchange slip, the Bank was warning of the possibility of an end to support, and depreciation of the speculative attack continued. It was not, however, a very skillful way of restoring confidence, and the reaction of the Banque de France made necessary the sterling pegging that continued until 20 September. On that date the cabinet prepared the legislation and the announcement abandoning the 1925 Gold Standard Act. In much greater secrecy, it also drew up a scheme for a banking holiday on the German model “in case any panic should occur.”108
The absence of a rise in the interest rate at which the Bank of England lent (Bank Rate) is also a puzzle. Although a 6 percent level was briefly considered, the Bank did not use one of the most traditional defenses of central banks. The Bank did its best after 5 August to minimize the drama of sterling, in order to protect British banks. Raising the rate would be an acknowledgment of the strain and an encouragement to get out while it was still possible and would allow the outward flow of short-term funds to continue. Raising the rate was also rejected because higher interest levels would send up the politically sensitive unemployment rate, and that might encourage a further speculative attack on the pound.109 Using reserves was likewise ruled out, since there was no point in doing this just to allow British banks to make payments and thus slide into illiquidity. Thus in practice the Bank of England did nothing to defend sterling.
Montagu Norman had broken down even before the pound sterling did. Faced with an apparently unstoppable speculative attack, Norman had on 28 July noted in his diary “feeling queer.” He then stayed in bed in complete nervous collapse until his doctors ordered him to go on a long sea voyage abroad. The Bank’s deputy governor informed him of the British devaluation in an apparently mystifying message which reached him on the ship home: “Old lady goes off on Monday.” When he reached the port in Liverpool, he received the deputy governor’s letter, which read: “Indeed we seem to have been afflicted with every kind of misfortune. I hope you are going to rest quietly for a bit.”110
The British bank dilemma made officials such as Harry Siepmann, the Bank’s major expert on international relations, seriously consider restrictions on capital movements as an alternative to devaluation. In addition, the more perceptive or suspicious Labour politicians became alarmed that the dispute about the dole was being made to carry the responsibility for a state of affairs that had nothing or little to do with the operation of the British labor market but arose out of financial conditions in central Europe. The secretary of state for the colonies, Sidney Webb (Lord Passfield), put the point in the following way: “If a foreign Government, which for reasons of internal politics was anxious to avoid introducing a system of national relief for the unemployed, chooses to take advantage of our exposed financial position, we can be held to ransom and compelled to make an essential change in our domestic social policy as the price of rescue for the financial interests of the City.”111
Interpreted in this light, the devaluation of sterling in September 1931 was a wholly successful operation. It did not remove the pressure for budgetary control. On the contrary, the national government won the general election on an austerity program designed to combat inflation, and continued to pursue balanced budgets. But it did halt the deposit loss; and deposits even increased as foreigners—and in particular Indians—exchanged gold for sterling.
This strategy was so successful because of the postdevaluation behavior of sterling on the foreign-exchange markets. If sterling had fallen only slightly, with an accompanying expectation of further falls, or if sterling had fallen continuously over a sustained period (as was the case with many currencies of South American countries after devaluation), investors would have remained nervous and continued to pull short-term funds out of British banks. However, the actual course was a sharp fall and then a stabilization, with a bounce back and some belief in a future rise (see Figure 2.2). This belief induced investors to leave their funds in London. In that way, devaluation as it was practiced in the British case saved the British banks. In that way, shaking off Britain’s “golden fetters” ended the British depression. It set the stage for a recovery based on a more relaxed monetary policy, which encouraged credit-driven spending on housing and consumer durables.
Figure 2.2 Pound and franc exchange rates relative to the U.S. dollar, 1924–1939
(Source: Global Financial Database, online at http://www/globalfindata.com, Exchange Rates)
After September 1931, financial pressure shifted to the United States. The story of the United States is clearly central to the whole story of the world depression. The United States was by far the world’s largest industrial economy, and of course the world’s largest capital exporter. A failure of its capital markets was a major part of the European and South American story.
The peak of American prosperity had been in 1929. But we should not assume that there was already a Great Depression with the stock market crash of October 1929. The downturn initially looked similar to—actually rather less intense than—the sharp postwar collapse of 1920–21. It took 1931, and a new sort of financial crisis, to turn the American experience into a Great Depression. The shock produced radically new, and very pessimistic, notions of the possibilities for public action in monetary and fiscal policy. Here, as elsewhere, the depression generated an intellectual paralysis.
The thesis that a contagious financial crisis had a major impact in worsening the U.S. depression may at first appear almost perverse. The mechanism for contagion seems very clear in the case of small central European or Latin American economies, dependent on the international capital market, and even in the case of Britain, where some strategically critical City banks had a major part of their assets invested in foreign undertakings. But in the United States, with its enormous internal market?
A powerful, and well-known, case has been made by Friedman and Schwartz that bank panics in the United States caused a dramatic contraction in the money supply that was not effectively countered by the Federal Reserve System. These panics, in tandem with the policy failure of the central bank, provided the monetary causes of the Great Depression. Friedman and Schwartz identified four waves of bank failures, November–December 1930, April 1931, September–October 1931, and finally February–March 1933. In the course of these, the number of banks in the United States was reduced from 24,026 at the end of 1929 to 14,440 by the end of 1933.
This analysis has stimulated a great deal of valuable research. Ben Bernanke has suggested a different, nonmonetary mechanism, whereby the failure of banks increased the cost of credit intermediation, and thus worsened general business conditions.112
As an explanation of the decline of the U.S. economy after 1929, the Friedman and Schwartz account is inadequate. In particular, as a recent detailed examination of the local circumstances of bank failures by Elmus Wicker has made clear, the first two of the Friedman and Schwartz waves of bank collapses have been misidentified. The first wave was not a general collapse, but rather a product of the failure of two institutions. The New York Bank of United States had $161 million in deposits. The Nashville Tennessee investment bank Caldwell & Company controlled the largest chain of banks in the South, with $200 million in assets, and its collapse in November 1930 had immediate effects in four southern states. The second wave was also a regional, rather than a national, phenomenon. In terms of Federal Reserve districts, it was limited to Chicago, Cleveland, and Philadelphia.
The most interesting, and most important, of the U.S. banking panics was the one immediately following the British departure from the gold standard. Unlike the previous localized panics, but like the European crises of 1931, this bank crisis was accompanied by an external drain. From 21 September to 8 October there was a gold outflow of $369 million from the United States, almost all to Europe, and the Federal Reserve Bank of New York (FRBNY) responded with the classic measure of an increase in its discount rate, from 1.5 to 2.5 percent.113 The governor of the New York bank, Harrison, on 8 October expressed doubt that an increase in the New York rate would really stop the outflow: elsewhere, such a measure had undermined rather than strengthened confidence. In the next week the gold outflow doubled, and on 15 October the discount rate was increased to 3.5 percent (still low in comparison with the extraordinarily high levels that central banks in central Europe had used in the failed defense of their currencies). Friedman and Schwartz comment that this external panic “was accompanied by a spectacular increase in bank failures and runs on banks.”114 Here, as elsewhere, their argument is somewhat slippery as to causality: did the external run affect the banks, as customers withdrew deposits in order to move out of the dollar? Or did doubtful customers of weak banks simply look for stronger institutions, and believe that they existed in Europe?
New York banks, which might be thought to be the most exposed to international problems and influences, were not in fact the worst affected by the September and October panics. As in the spring of 1931, the crisis was worst in the Midwest. Another round of this sort of attack resulting from foreign withdrawals occurred in the spring of 1932. The New York banks were the most conspicuous victims, as liabilities of U.S. banks to Europeans fell by $550 million. The Europeans withdrew deposits across the exchange, with the result that the U.S. monetary gold stock fell by almost exactly the same amount ($535 million).115
In response to the appearance of a crisis structurally similar to that of the European trauma of 1931, policymakers believed that they had no choice except to respond to the psychology, irrational as it might be, of the market. There was no room for fiscal maneuver, not because of any economic analysis of the consequences of larger deficits, but because of the (quite reasonable) belief that nervous markets would immediately punish fiscal deviancy. The same psychology explains why the Federal Reserve was so reluctant to pursue the monetary expansion—via open-market securities purchases—that Friedman and Schwartz reasonably believe might have stabilized the monetary situation.
Up to the September crisis, President Hoover had tried to deal with the crisis by offering bold, expansionary fiscal measures (which rather belie his usual reputation as the ineffective depression-era predecessor of the more imaginative Roosevelt). On 2 June 1931 Hoover had announced that the deficit for the fiscal year 1931 would exceed $900 million. This produced some hostile comments in the press, but until the sterling crisis the federal government found no problem selling treasury bills or certificates of indebtedness (indeed they are alleged to have gone “like hot cakes”).116
Then the impact of the sterling crisis led to a similar focus on public finance as had occurred in central Europe, and then in Britain. By the late autumn, the U.S. treasury bill situation swung round completely, and sales now affected prices in a weakened market. This in turn affected banks that held assets in securities. In December 1931 the Committee on Progress of Public Works of the President’s Organization on Unemployment Relief reported the new situation. Major issues of government bonds “would cause serious declines in the market values of the present outstanding low-yield issues, and thus result in severe losses for the holders of such securities. It may well be that one result would be a considerable number of additional bank failures.”117
By this time, maintaining—and preferably improving—bond prices became a major first step in any program to tackle depression. In part confidence could be manipulated by shifting as many items as possible off-budget: the new Reconstruction Finance Corporation of 1932, designed to support the banks, was allowed to borrow $1.5 billion in its own name, so that this would not appear as a part of the federal deficit.
Above all, a major tax increase, despite the procyclical consequences for demand, appeared the best way to reassure a nervous market. In his State of the Union message on 8 December 1931, Hoover took a sharply different line from that of only a few months earlier:
Our first step towards recovery is to reestablish confidence and thus restore the flow of credit which is the very basis of our economic life. The first requirement of confidence and of economic recovery is financial stability of the United States Government … Togofurther than these limits in either expenditures, taxes, or borrowing will destroy confidence, denude commerce and industry of their resources, jeopardize the financial system, and actually extend unemployment and demoralize agriculture rather than relieve it.118
The tax bill, and its success, now became the signal for success or failure in Hoover’s attempt to resist inflationism.
The New York Federal Reserve Bank blamed the precarious exchange situation on conflicts in Congress, which failed to pass the tax bill because the initial version contained a highly unpopular sales tax.119 On 3 May 1932 the U.S. Treasury announced that the deficit for the first ten months of the fiscal year had been $2.234 billion (by comparison, the previous year’s figure was merely $886 million). Only at the beginning of June did the Senate pass a budget-balancing bill raising income tax, excise, and postal rates in order to yield an additional $1.121 billion, and making economies (including pay cuts for public employees) of $238 million.
Foreign investors saw in the United States the budget problems that they had witnessed the previous year in central Europe. The U.S. position, in the English view, “was very shocking … a lot of serious talk was going on about the dollar and our leaving the gold standard.” The cause was not so much the bank failures but “politics in Congress and failure to balance the budget, which was taken by many people as evidence of a wish to get off the gold standard.”120 Governor Harrison of the New York Fed immediately used this information from London to press Secretary of the Treasury Ogden Mills and President Hoover to pass a tax bill and agree with Congress on an economy drive—in short, to perform all the deflationary routines that the U.S. bankers had forced on Europe over the previous year. Hoover’s only objection to this was that he could not act out of the blue and that he needed a crisis to force Congress to accept this position: “they needed some shock, like a wide open break in the stock market or the exchanges, for instance, to prompt immediate action.”121
The U.S. elections in November, in the campaign for which the Democratic candidate promised effective action to balance the federal budget, were accompanied by gold inflows: $16 million came in the week ending 21 November, $29 million the next week; and the movement out did not begin again until early February, when it coincided with another banking panic. From December 1932 to 15 March 1933, 447 banks were suspended, merged, or liquidated.122 Between 1 February and 4 March the FRBNY lost 61 percent of its gold reserves, and at the beginning of March the foreign-exchange market in dollars collapsed because of heavy selling by the British Exchange Equalization Fund.123 A recent analysis suggests that external conditions—the fear of devaluation—contributed significantly to the domestic panic.124 (An alternative account sees the final and most destructive of the U.S. banking panics as a consequence of the disorderly declaration of bank holidays in individual states, beginning in Michigan. Depositors in other states saw what was happening and tried to withdraw their funds before the declaration of new holidays.)125 The most plausible explanations of the two major general waves of bank failures (the only generalized panics in the United States) place the emphasis on foreign factors.
There were plenty of initiatives for dollar devaluation. Senator Thomas T. (Tom) Connally, a Texas Democrat, had tried to include this measure in the Glass-Steagall bill on reform of the monetary and banking system. There were signs that the new administration might be sympathetic. President-elect Franklin D. Roosevelt had held a meeting in December 1932 in Albany in order to discuss devaluation, and on 19 February Senator Carter Glass refused to become secretary of the Treasury because he had not been able to extract from Roosevelt a promise not to devalue. The national Banking Holiday eventually enacted in the early morning of 7 March 1933 became necessary in order to protect the Federal Reserve System from further losses, though naturally the Board wished to keep this a secret, since, as Chairman Eugene Meyer said, he “didn’t want the Federal Reserve System to be blamed.”126
The governor of the Federal Reserve Bank of New York, George Harrison, had since 1930 been an advocate of open-market purchases of government securities as a way of halting the dramatic monetary contraction. In April 1932 the program was stepped up, partly in order to forestall government legislation affecting the Federal Reserve System. By June 1932 total purchases amounted to $1 billion, but this only offset a loss of $500 million in gold and a reduction of $400 million in Federal Reserve discounts and bills bought. But given the extent of foreign nervousness, greater activity by the Federal Reserve System in this regard would have had a counterproductive effect.
Likewise, government initiatives against the crisis were effective to only a very limited extent. Hoover blamed enemies on two fronts for his dilemma. On the one hand, the bankers hemmed him in and had made the bold anticyclical schemes of the 1920s, or even of 1930, impossible to realize. Leading bankers, he said, quoting “one of the best and most influential” citizens, were “nothing less than public enemies, and all should be treated as such.”127 On the other side, Congress had damaged the government’s credit by debating freely on “wild” and “vast” schemes, and had upset the market. Squeezed between politics and the market, Hoover—and indeed every other statesman of the time—was completely and hopelessly lost.
Hoover’s attempts to regulate the financial markets in order to resolve the dilemma were unsuccessful. The National Credit Corporation, set up at Hoover’s insistence by the bankers in October 1931, was a private corporation intended to give loans to individual banks. The Reconstruction Finance Corporation of January 1932, a government body, gave $900 million in loans through 1932. The Glass-Steagall Act (February 1932) allowed member banks to borrow on more generous terms from the Federal Reserve System.
All these innovations failed to prevent the large wave of panics and banking failures after January 1933, which led to the need to impose a nationwide banking holiday. In part, the disaster of 1933 was simply a consequence of the mishandling of two large Detroit banks. Wicker believes that “if there had been bold and effective action initially to reopen the Michigan banks, the collapse of the banking system may have been averted.”128 The only person who might have had enough political power to step in was Hoover, but he was in the last weeks of his presidency, a lame duck in the original political sense of the term. It was the new administration that thus needed to deal with a double problem of internal and external runs.
Roosevelt in his election campaign had promised fiscal orthodoxy, in line with the requirements of confidence. His program included a 25 percent reduction in federal expenditure and a balanced budget: “I regard reduction in Federal spending as one of the most important issues of this campaign. It is the most direct and effective contribution the government can make to business.”129 This was a sincere belief, and not simply a tactical move to reassure a worried electorate. Until the deep depression of 1937–38 forced a rethinking, budget balancing played an important part in the formulation of fiscal policy. The only effective measure against the financial contagion was the dollar devaluation, which Roosevelt accepted on 18 April 1933. It was made even more effective by the announcement of 3 July 1933 that the United States would not attempt to stabilize the dollar parity. The message included the claim: “The world will not long be lulled by the specious fallacy of achieving a temporary and probably an artificial stability in foreign exchange on the part of a few large countries only.”130 Roosevelt’s measures made a major contribution to ending the bank runs, and with them the U.S. depression.
The effects of the contagious banking crises had included not only a direct and devastating effect on economic activity, but also a perception of the limitation on the room for fiscal maneuver by the federal government. Under Hoover, federal expenditures as a share of GNP had effectively doubled, from around 4 to 8 percent; but they could not go higher, because of the way in which the confidence limitation was perceived.
On an international level, the focus of speculation shifted elsewhere. France had received a substantial amount of international capital flight money, from central Europe in the first half of 1931, and then from Britain and especially the United States. The U.S. flows continued into 1933, so that an estimate of September 1933 gives the figure for capital flight into France as $8.3 billion, “of which a very large amount is American money,” fleeing from inflation and Roosevelt; in addition the French Finance Ministry estimated that 1.5 billion treasury bills (bons du trésor and bons de la défense nationale) were held abroad.131
This situation made France enormously vulnerable: should confidence return elsewhere, or should the French economy show any signs of inflation, the gigantic inflows would flood out again. The capital inflows put as much pressure on governments to maintain a deflationary course as fear of capital outflows did elsewhere. The major problem facing policymakers lay in the volatility of “hot money.” When did the first signs of a franc weakness become apparent? Already, it seems, before the U.S. dollar left gold.
An internal gold drain took place in late 1931 as French bankers and individuals feared inflation; and in particular those well-placed in financial circles took advantage of their information about the likely threats that followed from France’s large capital imports. According to the Bank of England, when the Banque de France’s gold losses were examined, “the names of two of the Regents of the Banque de France, one in his own name and one through his company, figured as purchasers of gold.”132 In December 1932 foreign banks were shipping gold out of France: the Guaranty Trust Company, the Banque Belge pour l’Etranger, and Barclays Bank.133 In the summer of 1932 the worry was, according to part of the French financial press, that France was in “pleine inflation.”134
But it was the U.S. gold embargo in April 1933 that sent France on a roller-coaster ride lasting for the rest of the decade. The gold exports of the United States to France halted and were reversed: in 1934 France sent $60.5 million to the United States, in 1935 $934.2 million, and in 1936 $573.7 million. The immediate reaction to the Roosevelt declaration, according to the Banque de France’s director of the Foreign Service, was that “London and most other European centers figured that France would leave the gold standard very shortly. The result was … that balances were being withdrawn, particularly to Switzerland.”135
The French in consequence were from 1933 highly concerned with their vulnerability to flows, and tried to deal with it by urging an international currency stabilization. But they were frustrated in the first place by the U.S. gold purchases of 1933–34 undertaken in order to drive down the dollar; and second by British unwillingness to commit sterling to a fixed parity.
The gold standard had always made necessary a difficult balancing act between internal and external monetary policies; and France had even before the central European crisis been plagued by budget and banking problems.
France rapidly departed from the large budget surpluses that followed the franc stabilization of 1926–1928. Large initial surpluses made it appear that there were plentiful funds available for new spending projects. By the budget year 1930–31 there was a small deficit, and it grew alarmingly throughout 1931. In addition, the depression, which came relatively late in France, began to affect revenues and expenditure. Industrial output fell from June 1930, though through 1931 the decline was relatively modest. The depression, and the question of how to adjust the budget, posed more and more of a strain on French political institutions.
In the 1932 election campaign, as in the United States, both sides claimed to represent fiscal probity and anti-inflationary orthodoxy. After the narrow victory of the cartel des gauches, Edouard Herriot’s government tried to demonstrate its concern for rectitude by appointing the technician Louis Germain-Martin as finance minister. He reduced the salaries of civil servants (fonctionnaires), cut expenditure, and pressed through tax increases. These measures were watered down after parliamentary debates, and the deficits remained. In December 1932 the cabinet fell over the war-debt payment due to the United States. In 1933 Edouard Daladier’s Radical government attempted a more rigorous implementation of budgetary deflation. Laws of 28 February, 31 May, and 23 December aimed at reducing the scope for tax evasion; and a supplementary levy was imposed on the fonctionnaires. In total, in 1932 and 1933, 5.612 billion francs were saved by cuts, and 3.15 billion gained by so-called tax adjustments.136 But the savings were canceled out by the government subsidization of agriculture required as the consequence of the imposition of a minimum grain price. As in central Europe, Germany, and also the United States, agricultural policy introduced a crucial element of destabilization into fiscal policy.
Parliamentary politics moved in two opposed directions: on the one hand, parties were unwilling to accept responsibility for cuts in payments—particularly to the fonctionnaires (who the socialists feared might be driven to support the Communist party). On the other hand, all parties from right to left had a substantial rural constituency, and pressed for greater spending on agricultural support. In a memorable phrase, Herriot spoke of the French peasant as the “greatest of French philosophers” and “our silent master.”137
At the end of 1933 the financial situation deteriorated, Daladier fell, and in February 1934 fascist demonstrations brought Paris to the edge of civil war. The public lost confidence in the government, but also lost confidence in the banks; and the latter loss made the problems of dealing with the budget more acute.
In November 1930 the first big bank failures occurred in France: the Banque Adam of Boulogne-sur-Mer and the Oustric group. The Oustric collapse raised questions about the Banque de France’s policy, judgment, and political contacts: it became a characteristically French financial “affair” with a parliamentary commission of inquiry that launched an onslaught on prominent politicians and on the governor of the Banque. The Banque was blamed for its overgenerous discounting of Oustric paper in the years before the collapse. It had raised the portfolio of Oustric bills from 6.66 million francs in September 1927 to 7.7 million in December, 20.385 million in April 1929, and 128.6 million by February 1930.138 The governor of the Banque de France, Clément Moret, tried to defend his actions in a rather high-handed way—by explaining that the Banque was a private institution whose autonomy was guaranteed by the state, and that he had no obligation to release the figures of the Oustric account.139 But Moret in the end dismissed the officials immediately responsible for the problem.
It was not just Oustric who had been heavily committed to investments in the rest of Europe. In August 1930 the four leading French banks had, the best estimates assumed, 5–6 billion francs abroad, of which two-thirds were invested via London; but 90 percent of the sum went eventually to central Europe, and London played no more than an intermediary role.140 In September 1930, after the German Reichstag elections, a substantial part of these French investments were repatriated.
Not only were there French investments—usually indirect—in central Europe; France also became a magnet for capital moving out of the financial trouble zones. One indication of the extent of the movement is given by the rise in private deposits at the Banque de France, which during the central European panic increased from 11.884 billion francs (27 March 1931) to 15.187 billion on 26 June 1931. The additional 3 billion represents a part of the movement of money out of central Europe.
At first the authorities tried to make light of the French difficulties. Governor Moret said that the French bank failures were “due less to intrinsically unsound positions than to a wave of exaggerated pessimism.”141 But speaking to the Regents soon after the first bank failures, Moret struck a rather gloomier note: “There is a psychological factor that entirely escapes the action of the Banque. Movements of capital are today determined less by differences in interest rates than by the greater or lesser security they offer. Now, in the troubled state of the world, in the presence of the worries that surface in many countries, the franc appears as one of the most solid currencies—as a currency of refuge.”142
Highly volatile capital flight movements led to new problems for France. Within a few years, anyone who wanted to restore confidence in French banking and in the French economy in general had to deny that there were large sums of foreign-flight capital placed there.143
The legacy of the Oustric affair was that on the one hand, the Banque de France suffered criticism for the overgenerous rescue of fraudulent enterprises and, on the other, bank failures prompted demands for a more expansive central bank policy. In 1931 Chambers of Commerce protested the “excessively restrictive discount and credit policy.” In 1932 the Regional Association of Chambers of Commerce in Grenoble noted a resolution of its members that “the Banque de France should demonstrate the greatest liberalism in the granting of rediscounts.”144
In fact the Banque de France did extend new credits to numerous French banks in the wake of the banking runs. During the first crisis (October 1930–January 1931) the Banque’s portfolio of bills (an indication of its lending) increased from 4.685 billion francs to 7.364 billion. It rose again in September 1931, in the aftermath of the sterling crisis and the failure of the sixth-largest French bank, the Banque Nationale de Crédit (BNC).145 The Banque de France tackled the panic by giving new credits to rescue the Nancéienne de Crédit, the Banque Privée, the Groupe Messine, and the Mines d’Anzin, as well as to the Société Générale.146
The initial crisis in France had been an affair of the second-rank banks. The Banque d’Alsace-Lorraine had been weakened by the repercussions of the German inflation of the early 1920s, and in addition had taken over a series of weak banks in the late 1920s, of which the largest was the Banque du Rhin. The BNC had been in a very shaky position in the early 1920s and had had heavy losses during the franc inflation; it had experienced only a brief recovery in 1926–1928, when it had been able to reduce its obligation to the Banque de France from 480 million francs to 3 million.147
From December 1932, however, all the major French banks lost deposits. As they became affected, they ran down their accounts with the Banque de France (over four months their holdings fell from 8 to 3 billion francs); and they asked for rediscounting at the Banque. They protested that they could no longer absorb the large volume of state paper issued.148 Foreign banks (notably the Morgans and Bankers Trust) began to discount at the Banque de France too,149 and a new series of provincial bank failures hit France: Charpentier (Cognac), Société Saint-Quentinoise de Crédit, the Banque du Centre (Limoges), Banque Renauld (Nancy).
In 1934 another major banking crisis threatened, and the big banks again looked vulnerable. There were difficulties at the Banque de l’Union Parisienne, the institution most concerned with the financing of the French armaments industry. The Crédit Lyonnais expanded its discounts with the Banque de France in February; and Paribas also needed the assistance of the central bank.150 From the end of 1932 to the end of 1936 deposits at the four great banks fell from 21 billion francs to 15.3 billion (while savings banks retained a greater degree of public confidence, and their deposits actually rose, from 34.216 to 35.714 billion francs).151
The movements after 1932 are an indication of the external drain afflicting France. When bank-deposit withdrawals threatened the French credit structure, the Banque de France increased its rediscounting in order to keep the French banks liquid; but this made resources available for outward movements, and in this way the Banque while propping up the French banking structure actually nourished the flight of capital from the country.
The process can be studied quite precisely in the big franc crises of the early 1930s. From 2 December 1932 to 31 March 1933 the Banque de France lost 3,354 million francs in gold; from 15 September to 22 December 1933, 6,774 million; from 12 January to 2 March 1934, 3.401 billion; from 29 March to 7 June 1935, 11.684 billion; and from 25 October to 6 December 1935, 6.198 billion francs. In all these crises, the Banque’s portfolio increased, by 719, 1,041, 1,646, 5,430, and 3,289 million francs respectively, in line with the external capital movements; and the credit account of the great banks contracted (2,013, 835, 736, 2,206, and 657 million francs respectively).152
By 1935 the gold outflows had assumed such alarming proportions that internally at least some officials of the Banque de France began to doubt the Banque’s official line of doing everything to resist franc devaluation. On 5 March, just before the first big panic wave of 1935 set in, the director of the Banque de France’s Foreign Department, Charles Cariguel, “thought that the majority of businessmen in France were willing to make the sacrifices necessary to adjust to something like the present value of sterling but if the pound depreciated further they would probably regard adjustment as hopeless and be unwilling to make further sacrifices to that end … For the first time during my telephone conversations with him, Mr. Cariguel clearly implied that devaluation of the franc might be necessary to adjust their position vis-à-vis sterling.”153
By April, Cariguel was filled with gloom about the franc, since the Belgian devaluation appeared to destroy the gold bloc; he predicted that Switzerland would follow rapidly, and then France would be exposed to the full gale. On 15 April, J. E. Crane, the deputy governor of the Federal Reserve Bank of New York, made a suggestion that sketched out the line that would eventually be followed in September 1936: he inquired about rumors of a 15–20 percent French franc devaluation: “that done, there was some possibility of a three-cornered arrangement—the dollar, the franc and sterling. He asked Cariguel how the French public would take that and Cariguel said that in his opinion the French were not ready for such a plan even if it were offered as part of an exchange stabilization all around.”154
There were bolder spirits than Cariguel. The first major public statement in support of devaluation had come in June 1934 in a speech to the Chamber of Deputies by the conservative politician Paul Reynaud (who had in private been convinced of the need to devalue since 1933). He faced a massive onslaught from the press, who vilified him as the would-be expropriator of French small savers. Reynaud found only one major press ally, the deputy and proprietor of Le petit journal, Raymond Pâtenotre. Professor Charles Rist, an influential economist and as a former deputy governor of the Banque de France something of an “insider,” pleaded for devaluation from the spring of 1935. So also, most influentially, did the French financial attaché in London, Emanuel Mönick. On 1 September 1935 Mönick concluded that Britain was not likely to stabilize, and that France resembled the United States in the last days of the Hoover administration. He deduced that a “wisely measured monetary adaptation”—a circumlocution for devaluation—was needed.155
By early 1935 France had reached political and economic deadlock. Failure to stabilize the French budget by the familiar measures—to cut back expenditure, slash civil service pay, and raise taxes—meant movements across the exchanges, bank withdrawals, and increased borrowing from the Banque de France. Money rates in Paris soared in consequence. Some of the leading figures at the Banque de France, such as Robert Lacour-Gayet, the director of the Service d’Etudes, thought that governmental self-discipline was all that was needed: “A show of determination to put their house in order could very rapidly restore confidence and put an end to the whole of the present movement based on fear and speculation.”156 In November the business world blamed a new run on the franc on agitation in the Chamber’s Commission des Finances for an end to the austerity policy.157
But France had reached by now the paradoxical position that the strains were so great that even deflation would undermine confidence, because of the heightened threat to the social order. Already the franc panic of early 1934 had been set off by demonstrations of the fascist Leagues in the streets of Paris on 6 February. These demonstrations had produced a mobilization on the left, and the formation of a Popular Front that included Communists. The center-right government lost support to both sides. The international market reacted by reading further deflation as a sign, not of an upgrade, but of a new menace to stability.158
Once France had reached this position, devaluation (which came, under a Popular Front left-coalition government, in October 1936) was the only option for breaking out of the vicious cycle of bank failures and imposed deflation.
France did not suffer alone in the 1930s. The other countries remaining on gold—the so-called gold bloc of Switzerland, the Netherlands, and Belgium—also suffered from chronic financial instability in which speculative movements played a major part.
These economies were vulnerable on two grounds. The ties to Germany of Switzerland and the Netherlands presented a grave problem. Both countries attracted a great deal of capital flight that might move out again rapidly. Moreover, Swiss and Dutch money was frozen into Germany under the standstill agreements. Short-term loans in the amount of 900 million Swiss francs loans were subject to the German standstill; but that was not the end of the Swiss problem. By 1934 a total of 3.39 billion in debts were subject to compensation agreements. In 1933 Dutch credits to Germany amounted to 1.044 billion guilders. In addition, German shares to the nominal value of 263.8 million guilders were held by Dutchmen.159
Second, the maintenance of convertibility at the old parities subjected these countries to speculative flows reacting to British, French, or American policies or anxieties. They had to pay an ever higher price for the maintenance of an open economy.
The international exchanges reveal the extent of Swiss volatility in the first half of the 1930s. Until the autumn of 1931 the Swiss Nationalbank gained substantial amounts of gold: by the beginning of September it had gained 1.158 billion Swiss francs (or 97 percent) in the year. From the sterling crisis until the devaluation of the dollar, its gold holdings remained steady. After this, there were crises that corresponded with problems in France. From March to July 1933 the Nationalbank lost 716 million Swiss francs in gold; from January to April 1934, 364 million; and from March to June 1935 517 million.
These flows in and out also affected the liquidity position of the Swiss banks. The big inflows of 1931 resulted in enormously bloated cash reserves in the commercial banks (these rose from 467 million Swiss francs in 1930 to 1.29 billion in 1931—in other words, they account for almost all of the gold inflow).
In Switzerland a major political debate centered upon the consequences. “Capital export,” its critics said, had brought few gains and benefited only a small urban financial elite, while exposing the country to external risks and random shocks from international politics. The peasant political leader Gottfried Gnägi argued that there should be no capital export as long as Swiss interest rates remained at crisis levels.160 The problem lay in the way in which external involvement undermined Swiss financial stability.
After the collapse of the summer of 1931, capital flows were regulated. A “gentlemen’s agreement” concluded by the banks in February 1932 involved a promise that they would consult the Nationalbank before embarking on foreign capital issues; and this regulation of long-term loan activity intensified during the decade. On the other hand, the Nationalbank had no control, and indeed no information, about short-term movements. When, in the course of trade negotiations, the government and the Nationalbank attempted to find out about the volume of Swiss assets in Germany, the Swiss banks simply refused to reply.161 In late 1934 a banking commission had been formed in order to investigate Swiss banking problems, and the report warned that the Swiss exposure to Germany presented great risks, and that even a partial guarantee of the German assets could not remove the likelihood of financial panics.
The banks’ obstinacy appeared self-defeating in view of their vulnerability and their dependence on government support; but it was also in reality a way of protecting themselves from revelations about just how weak they really were. In order to mask their difficulties the Swiss banks insisted that bank secrecy (Article 9 of the interwar Bank Law) formed an essential element of the Swiss way of life.162
In fact by 1934 a substantial amount of damage had already occurred. One of the eight great banks, the Geneva Banque de l’Escompte, had already been in trouble in 1931. In 1932 it reported that of its total assets of 392 million Swiss francs, 200 million were frozen in foreign investments (mostly in Hungary and Germany), and that it would be forced to declare bankruptcy if Germany imposed a moratorium.163 The state supported the ailing bank, lending 55 million Swiss francs through a newly created government-owned loan institute (the Schweizerische Eidgenössische Darlehnskasse), and participated to the extent of 20 million Swiss francs in a recapitalization of the bank.164 But even this reconstruction involved an optimistic valuation of the assets and a willful ignoring of the severity of the central European crisis and its impact on the Swiss. At the government discussions about assistance to the Banque de l’Escompte, one banker explained: “In valuing the assets we can’t use the principle of a bon père de famille. That is possible only in normal times.” But there existed little doubt about the urgency of the task. In the view of Federal Councillor Jean-Marie Musy, “A sudden closure of the Banque de l’Escompte would have an influence on the stability of our currency. The higher interests of our country demand that we intervene to save the bank.”165
In November 1933 another of the great banks, the Banque Populaire Suisse (Schweizerische Volksbank), also needed reorganization after 150 million Swiss franc credits were written off. And in the wake of this 1933 crisis, the Swiss commercial banks lost deposits, and there was a movement toward the cantonal banks (which were less exposed to foreign risks).
The German and central European connection did not represent the only source of instability in Swiss financial life. The bank commission echoed the complaints made in other countries about the existence of unbalanced public budgets; in the Swiss case, the cantons suffered especially from increased social obligations during the depression even as their revenues were collapsing. But—as in other countries—it proved convenient to blame outsiders and foreigners.
The persistent deflationary pressure on public spending—since even small deficits could provoke dramatic flights of money in and out—combined with a tradition of populist democracy, eventually provoked a revolt. On the one side, the peasant party argued against Swiss financial dependence on other countries; on the other, the socialists organized a legislative initiative calling for a program to combat the depression. It included a detailed set of work-creation measures. At the same time, government funds would provide debt relief, promote tourism, and stimulate exports. The program would be accompanied by the imposition of control over financial markets and the prohibition of capital export. The total cost of this ambitious scheme amounted to around a third of national income. Its opponents argued that it would bring at worst panic and best devaluation of the Swiss franc, and on the strength of this objection the initiative was narrowly defeated in June 1935. It was a very bitter debate that deeply split Swiss politics. Here was a case in which it might be convenient to blame the foreigner for the destabilization of Swiss affairs.
International banking provided a convenient scapegoat. The Swiss foreign ministry rightly believed that it was English merchant banks (not the clearing banks) that had suffered in 1931. The diplomats argued that the City of London hoped that international confidence in sterling might return if other countries plunged over the abyss, and as a result spread rumors about the instability of the gold-bloc countries. The maverick British Financial News correspondent Paul Einzig had indeed given wide publicity to stories about a massive Swiss and Dutch capital flight. But this was not the fundamental cause of Swiss difficulties. Given the international position, the only way of safeguarding Switzerland from dramatic exchange movements lay in an alteration of parity.
The first country of the gold bloc to succumb to the strains of being a small and relatively open economy on gold was Belgium. Once again the immediate impetus for devaluation was not a theoretical plan or the convincing arguments of economists, but rather a banking crisis.166 Belgian banks stood firmly in the continental tradition of universal banking, and in consequence faced heavy losses on their industrial lending and participations. The bank losses prompted an external drain in August 1934, and by March 1935 all the Belgian banks except the largest (the Société Générale) were close to collapse. Industry protested against attempts to rescue the banks and prevent further capital outflows by raising Belgian interest rates, since this, it claimed, would only intensify the business depression.167 Devaluation was thus the only option (March), and the pressure on the Netherlands and Switzerland increased.
In the Netherlands, with less of a tradition of universal banking and no threat of a major banking collapse, the government of Hendrik Colijn followed a course analogous to that of Heinrich Brüning in Germany or Pierre Laval in France. Systematic deflation, in which government spending, wages, and prices were all cut, offered a way of spreading the costs that seemed equitable. Some of Colijn’s supporters even cited the “success” of Brüning’s policy as evidence for the desirability of a consistent deflation,168 and Colijn himself had built his political reputation on his successful battle against inflation in the early 1920s. He thus resisted devaluation strenuously.
In the September 1936 crisis, French developments affected the smaller gold-bloc states. In addition, the Swiss market had been disturbed by the flotation on 21 September of a very large defense loan. The Swiss now lost gold at a fast rate—$9.5 million on 26 September alone—and quickly followed the French franc by devaluing in the range of 25.94–34.55 percent. Unlike in France, the devaluation was an unambiguous success. It offered above all a way of establishing social peace in Switzerland. The 1937 labor agreement, which set a highly managed and corporatist framework for the resolution of conflicts about wages and conditions, would have been impossible without the greater confidence that followed the ending of the large capital movements.169
Once Switzerland had left, Colijn was persuaded by the governor of the central bank that the Netherlands could not stand the strain of being the world’s only country on gold at the old parity. The Dutch followed France’s example and devalued to a new range of gold values, between 20 and 25 percent below the previous level. Escaping from the gold bloc proved yet again to be the only way of avoiding banking instability.
The gold standard had provided a transmission mechanism that made economies with weak banking systems vulnerable to shock. Central banks, which had originally been conceived as an institutional mechanism for alleviating the domestic consequences of external monetary shocks, had increased the instability because of the prevalence of a concern for preserving confidence. The central European weakness was especially acute because of the destruction of capital in postwar hyperinflations, and the weakened banks provided a poor basis for subsequent borrowing.
In order to find counterexamples, in which banks could withstand the strains imposed by the international financial system, it is necessary to move a long way from the world of the Genoa conference and the Geneva ideology of internationalism. Both Canada and Japan proved financially quite robust; and such robustness shortened the depression and provided a potential for successful recovery policies.
The order of events in Japan’s commitment to the international economy was quite different from the European picture of stabilization on the basis of a gold-standard parity (nominal anchor), capital inflows, loss of confidence, and banking and fiscal crisis. Japan experienced capital inflows, and then a banking crisis in 1927, prior to a commitment to the gold standard. The major capital inflows occurred before 1927. In April 1927 domestic bank runs set in.
Before the 1927 crisis, the governor of the Bank of Japan, Inoue Junnosoke, who fully supported the new international vision of the responsibilities of central banks in a fixed-parity world, had been a major advocate of a return to gold, which he saw as the answer to Japan’s high inflation and interest rates. But this course had been opposed by many major figures in the banking establishment (Yashiro Norihiko, of Sumitomo Bank; Kusihida Manzo, of Mitsubishi Bank; Ikeda Kenzo, of Daihyakyu Bank; Kodama Kenji, of the Yokohama Specie Bank).170 The Tokyo earthquake of 1923 further hindered the idea of a return to fixed parity. After the bank crisis of 1927, a wave of bank mergers strengthened the banking system, and some former opponents of the gold standard began to be sympathetic to the idea of stabilization. Inoue’s power increased, and in July 1929 he became finance minister. Just six months later, at the beginning of 1930, he announced Japan’s return to a gold parity.
Japan went off the gold standard again quite quickly, at the end of 1931, not because of the kind of banking and financial panic that afflicted central Europe and then Britain, the United States, and finally the gold bloc, but because of the trade effects of the British devaluation in September 1931. At the same time, a political change brought in a party opposed to gold (again, a novelty in comparison with other countries, in which the attachment to gold was generally part of the accoutrement of political respectability as surely as was the wearing of a hat). The Seiyukai party came to power in a disreputable way in December 1931, as the result of military pressure (the new government is sometimes described as the product of a coup). Only one day after his appointment, the new prime minister, Inukai Tsuyoshi, took Japan off the gold standard, and within two months won a major election victory.
The basis for a sustained recovery was laid by a foreign-exchange policy that let the yen depreciate and generated a revival in exports, especially of cotton textiles. Average annual growth for the recovery years 1931–1936 was 4.3 percent—an impressive contrast with the depression-ridden world elsewhere. The export offensive was accompanied by increased militarization, with larger public spending, raising overall demand in the Japanese economy. From 1932 to 1935 government expenditure on both emergency relief and military expansion following the 1931 Manchuria Incident was financed largely through bond issues, of which four-fifths were refinanced through the Bank of Japan. Deficits amounted to about 30 percent of government expenditure. Only in 1935 did the expansionary finance minister Takahashi Korekiyo attempt to rein in the inflationary deficits and adopt a policy known as “the gradual rationing of government bond issues.” But this application of economic brakes quickly produced a military mutiny, and Takahashi was assassinated in February 1936. Military expenditure subsequently increased even more rapidly.171
Canada had an equally happy economic (if less politically turbulent) experience, although it remained on the gold standard for a substantial time. The Canadian strength lay in a more robust banking system than that of the southern neighbor. Branch banking was less vulnerable than single banks or statewide banks to regional weaknesses, and there were—surprisingly—no Canadian bank failures despite the massive vulnerability of the economy (including the financial sector) to problems caused by the fall in the price of wheat, Canada’s major export. Nevertheless, the decline in Canadian real GDP and the rise in unemployment were similar to the U.S. experience: the depression was transmitted not through the financial system but through trade.172
A second strength lay in the fact that Canada, which then had no central bank, was consequently not vulnerable to the kind of institutional shock that had affected other countries as a result of the restrictive actions of central banks.173
An initial response to the economic turmoil of the 1930s involved the abandonment of all the elements of the Genoa-Geneva consensus. The gold standard was discredited. Orthodox fiscal policy, which was required in order to generate gold-standard credibility, was discredited. The heroes of the 1930s were expansionists, who managed to find new ways of financing state deficits: Finance Minister Takahashi in Japan or Economics Minister Hjalmar Schacht in Germany.
Also discredited were independent or autonomous central banks, which had been so crucial to the vision of how to restore economic order in the 1920s. And so too were capital movements.
Central banks engaged in a curious rearguard action, which may be responsible for much continued intellectual confusion, when they defended themselves by insisting that monetary policy had no great impact on economic activity.
The organ of central bank cooperation, the BIS, transformed itself into an institution for economic analysis, in brilliantly conceived annual reports from the distinguished pen of the Swedish economist Per Jacobsson, and for the collection of statistics about the world economy. None of this impressed Montagu Norman, who had a quite different concept of what was involved in central bank cooperation. In December 1932 he told a meeting of central bank governors at the BIS that he was “against statistics: he thought the figures were misleading and he believed that if central banks or currency Authorities worked on statistics, even the best statistics, they were more likely to be misled than anything else.”174
This collapse of the BIS into a center for merely routine operations was part of a broader breakdown of the theory of central bank action. As the depression deepened, and as criticism mounted on all sides, central bankers more and more believed that their only mission lay in announcing loudly that they could do nothing, that monetary policy could not influence the development of the real economy. This was a complete break with the central bank activism of the mid-1920s. It was also of course theoretical nonsense, which arose out of the (forgivable) feeling that politicians’ rather than central bankers’ blunders had made the financial mess. If this was what was meant by a “common body of monetary doctrine,” it was one that led away from giving central banks great room for maneuver in international financial matters.
In dealing with the League of Nations’ inquiry into the gold problem, the central bankers adopted the position that monetary policy was ineffective, and their view informed the majority report of June 1932 (a more far-ranging minority report, signed by Sir Henry Strakosch as well as Sir Reginald Mant and Albert Janssen, recommended concerted international action to raise commodity prices).
The modest recommendations centered upon the restoration of freedom of exchange. Central banks should allow the automatism of the gold standard to operate: “gold movements must not be prevented from making their influence felt both in the country losing gold and in the country receiving gold.” Governments were to take the burden of adjustment, accumulating budget surpluses and repaying debt in the deficit countries:
in each individual country the necessary steps shall be taken to restore and to maintain equilibrium in the national economy. This means that the budgets of the State and other public bodies must be balanced on sound principles, and also that the national economic system as a whole, and especially costs of production and costs of living, should be adjusted to the international economic and financial position, so as to enable the country to restore or to maintain the equilibrium of its balance of international payments.175
In private, the central bank consensus was stated even more explicitly:
We are quite unwilling to lend our authority to those who would exonerate politicians and businessmen from responsibility by explaining the terrible tragedy of the present world crisis as being due solely to a scarcity of gold … But it was evident to the Delegation, as is clearly expressed in the Second Report, that the causes responsible for this maldistribution were mainly of a general economic, financial and political nature. As these causes were not primarily monetary, monetary policies could not be expected to cure the world of the resulting ills.
The Italian finance minister Guido Jung explained that “it would be disastrous to the reconstruction of the world if in a report of ours we were to give to people the impression that there exists a monetary witchcraft, which can, by its own force, work miracles and avoid the necessity of … solving the political and economic problems, which have brought the world to its present conditions.”176
Part of the task of the 1933 World Economic Conference in London lay in discussing the contribution of central banking policy to crisis, but the central bankers themselves resented the interference. The preliminary meeting of American economic experts held at the Federal Reserve Bank of New York was quite characteristic. In the presence of Hoover’s secretary of state (Henry Stimson) and secretary of the Treasury (Ogden Mills), Governor Harrison and Chairman of the Board Eugene Meyer “emphasized the necessity of keeping off the agenda of a governmental conference purely central bank questions such as for example central bank credit and gold policies. They also pointed out that most of the monetary questions which could be placed on the agenda were of interest to central banks and that they thought it was of the utmost importance for the World Conference to avoid invading the central bank field or making any suggestions or giving any instructions to central banks which might prove embarrassing.”177
At an unofficial meeting of the BIS governors in February 1933, the president of the Belgian national bank urged against any central bank agreement before London, because this “would give a catastrophic reinforcement to the erroneous idea that the monetary factor is a primary factor which plays a preponderant role in the world crisis.” Norman agreed wholeheartedly. Eventually the BIS governors did produce a document to preempt London, titled “Rules of the Gold Standard.” It contained, perhaps it is needless to say, nothing but platitudes.
Central banks, orthodox finance, and international capital movements were such obvious villains in the drama of the depression that few commentators at the time bothered to devote attention to the problems of financial-sector instability. Indeed this aspect of the problems of the interwar order received more attention only with the rapid globalization of financial markets since the 1970s, in the course of which some of the problems of the older order began to reemerge.
Recent literature suggests that in many cases since the 1970s, banking difficulties preceded currency or foreign-exchange crises. It is easy to see some of the mechanisms: central banks or governments try to deal with actual or even incipient banking difficulties by monetary expansion or by fiscal actions (socializing bank debt, as in Mexico in the 1990s). The result is a real overvaluation, which creates the potential for speculative attack on the currency. This literature also frequently suggests that the banking crises are preceded by financial liberalization and deregulation, which encourages potentially destabilizing capital inflows. If these arguments were already being formulated as a response to the problems of the European Monetary System and Mexican crises of the early 1990s, the effect of the Asian crisis after 1997 strengthened this line of interpretation.178
The experience of the first stage of the Great Depression offers an analogy with this very contemporary problem. Banking instability played a critical part in creating the potential for currency crises. In the interwar period, the problem lay in central Europe in the tradition of universal banking, reconstructed in the wake of an inflationary shock on an undercapitalized base, and in the United States in the poor development of branch banking. In the 1990s, in East Asia, many Latin American countries, and Russia, the vulnerability lies in poorly developed accounting systems, corruption, and the intrusion of politics. In both periods thin markets, which dried up in a panic, made the weaknesses worse. The critical question for today is the extent to which the financial exposure of industrial countries may lead to a repetition of the story of the second half of the Great Depression experience.
One way of presenting the case for alarm about the transfer of crisis from the periphery to the core of the international financial system is in the form of a table. It shows, for various years and countries, banks’ gross foreign liabilities as a share of international reserves (see Table 2.4). A brief inspection will confirm some of the themes discussed above: countries are vulnerable when their banks have a high share of foreign liabilities relative to reserves, and the precursors of the outbreak of the crisis are dramatic changes in the preceding year or years. The G-7 figures may give an indication of a potential vulnerability, at least in the economies in which the financial sector plays a large part. The United Kingdom, as in 1931, is the most vulnerable to this kind of problem.
Ultimately, however, schematic overviews are not very useful as accurate predictors of likely sources of financial and banking weakness. There were few anticipatory exposés of the problems of Korean and Thai banks before 1997, just as in the United Kingdom an official report (by the Macmillan Committee on Trade and Industry) celebrating the superiority of the German banking system over the British was by chance published on precisely the day, 13 July 1931, on which Germany’s most famous risk-taking industrial bank failed and brought down with it the entire German banking system.
| Country | Share | % annual change |
| Germany (July 1931) | 3.7 | –25 |
| Mexico (3d quarter 1994) | 0.6 | 33.7 |
| East Asia (mid-1997) | ||
| South Korea | 1.3 | 15.1 |
| Indonesia | 0.7 | –10.1 |
| Thailand | 1.7 | 17.8 |
| Singapore | 0.8 | 5.9 |
| Hong Kong | 9.4 | –13.5 |
| Philippines | 1.7 | 39.1 |
| China | 0.5 | –21.3 |
| G-7 (1st quarter 1999) | ||
| Germany | 7.2 | –5.3 |
| United Kingdom | 64.7 | 12.3 |
Sources: Harold James, The German Slump: Politics and Economics, 1924–1936 (Oxford: Oxford University Press, 1986), p. 294; and IMF, International Financial Statistics, various issues.
Boom-bust cycles of international credit had of course been a characteristic feature of the nineteenth-century globalized economy, with many failures in the periphery. Was 1930–31 different in that it was a systemic crisis of the gold-standard system? In the nineteenth century the gold standard had been robust enough that the countries at the core of the system in western Europe never contemplated a departure from gold, even though the United states in the mid-1890s came close to being pushed off. The 1920s at first apparently followed the course of a typically nineteenth-century credit boom, but the bust in 1930–31 then assumed a systemic character that destroyed the gold-exchange standard. The international financial system in the 1990s proved more robust than some commentators suggested and feared. But there are vulnerabilities, which resemble the interwar weaknesses. Fixed exchange-rate commitments represent an easy transmission mechanism for crisis, and although they are now unpopular with economists, they continue to be widely regarded as the only realistic solution to obtaining capital in many mid-income emerging markets.
The preceding survey of banks and financial panics in the depression era and any reflection on the contrasts with 1997–98 also make clear the centrality of appropriate monetary policy in the major economies. The response in October 1998 to the threat of global financial contagion was a monetary easing, notably in the United States, combined with a dedication of additional resources to international financial institutions. Under the impact of the Brazil crisis and its consequences for the rest of Latin America, the U.S. Congress finally accepted a quota increase for the IMF, which previously it had resisted.
But the major lesson of both the 1920s and the 1990s experiences is that financial-sector stability is a key element in preventing self-destructive panics in a world of globalized capital.