4259 Prels 4/12/02 9:41 am Page vii
Chapter Title
1 2 3 4 5 6 7 8 9 10111 11 12 13 14 15 16 17 18 19 20111 21 22 23 24...
48 downloads
1895 Views
17MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
4259 Prels 4/12/02 9:41 am Page vii
Chapter Title
1 2 3 4 5 6 7 8 9 10111 11 12 13 14 15 16 17 18 19 20111 21 22 23 24 25 26 27 28 29 30111 31 32 33 34 35 36 37 38 39 40111
vii
LIST OF CONTRIBUTORS Thomas Ferguson
Department of Political Science, University of Massachusetts, Boston, USA
David Greasley
Dapartment of Economic History, Universty of Edinburgh, U.K.
Jakob B. Madsen
Department of Economics and Finance, Brunel University, U.K.
John H. Munro
Department of Economics, University of Toronto, Canada
Ronnie J. Phillips
Department of Economics, Colorado State University, USA
Gary Richardson
Department of Economics, University of California, Irvine, USA
Andrew J. Seltzer
Department of Economics, Royal Holloway College, Surrey, U.K.
Fred H. Smith
Department of Economics, Davidson College, North Carolina, USA
Peter Temin
Department of Economics, Massachusetts Institute of Technology, USA
Simone A. Wegge
Department of Political Science, Economics, and Philosophy, College of Staten Island – CUNY, New York, USA vii
4259 Prels 4/12/02 9:41 am Page ix
1 2 3 4 5 6 7 8 9 10111 11 12 13 14 15 16 17 18 19 20111 21 22 23 24 25 26 27 28 29 30111 31 32 33 34 35 36 37 38 39 40111
INTRODUCTION Volume 21 of Research in Economic History is a substantial contribution in several respects. Its heft reflects the continuing increase in quality submissions to this series, which invites (although it does not require) authors to take advantage of less stringent space limitations than is typically true in a journal article. In different ways, this opportunity has been exploited by John Munro, in his encyclopedic treatment of sticky wages in Medieval England and the Low Countries, by Gary Richardson in his systematic reconsideration of Deirdre McCloskey’s explanation of scattering among medieval peasants, by Thomas Ferguson and Peter Temin in their controversial reinterpretation of the 1931 currency crisis in Germany, by Simone Wegge in her exploration of the determinants of immigration from the German principality of Hesse-Cassell, and by Fred Smith in his painstaking quantitative investigation of the history and geography of land values in the city of Cleveland. In somewhat shorter but no less substantial contributions, Ronnie J. Phillips uses the records of the San Francisco Clearinghouse to study how banking services persisted in the face of the 1906 Earthquake, and Andrew Seltzer examines personnel records from the Union bank in Australia in 1931 to determine the actual effect of a mandated 10% nominal wage cut. Finally, David Greasley and Jakob Madsen revisit econometrically the causes of the precipitous drop in consumption spending in the United States between 1930 and 1932 (this discussion is not necessarily in the order in which the articles appear in the volume). The papers offer regional diversity: two with principal focus on England, one on Germany, one on Australia, and three on the United States. There are some commonalities in themes: we have three papers on 1931, three papers that have something to do with banks, two on urban economic history, and two on wage stickiness, albeit in different countries and addressing labor markets several centuries apart. What can be said of all of these inquiries, however, is that each involves the careful consideration of quantitative and qualitative data within a well articulated theoretical framework. And in almost every case, we have original analysis of primary source material. The volume begins with Thomas Ferguson and Peter Temin’s provocative essay on the German currency crisis of 1931. Marshalling an impressive array of archival and statistical material, they challenge existing interpretations which have focused on a banking crisis and argue instead that this was a run on the ix
4259 Prels 4/12/02 9:41 am Page x
x
1 2 3 4 5 6 7 8 9 10111 11 12 13 14 15 16 17 18 19 20111 21 22 23 24 25 26 27 28 29 30111 31 32 33 34 35 36 37 38 39 40111
INTRODUCTION
currency, caused proximately by ill-advised budgetary policy and by a shortsighted foreign policy that cut the ground out from under the possibility of French assistance. In her study of nineteenth century immigration from the German principality of Hesse-Cassel, Simone Wegge links data on individuals to socioeconomic characteristics of their villages of origin. Wegge’s work is based on original archival work. She has carefully linked data on individuals from emigration lists with census data for the same individuals and village-level data. The result is an important addition to our knowledge of the determinants of nineteenth century European migration. With a similar attention to primary source material, Fred Smith develops a remarkable economic history of the city of Cleveland, exploiting assessors’ data to analyze its changing fortunes by charting the fluctuations of land values across space and time. His most striking finding is that the bulk of the substantial deterioration in land values occurring between 1950 and 1980 took place in the 1970s, and does not seem easily explicable as the consequence of the changing demographic characteristics of neighborhoods. Ronnie J. Phillips explores the response of the San Francisco banking community to the 1906 Earthquake, enriching his narrative with quotations from the minutes of the committee of the San Francisco Clearinghouse. He offers us a fascinating case study of a private financial system’s response to disaster, focusing on how bank runs were avoided through a combination of collaboration within the private sector as well as between banks and the U.S. Mint. An Appendix compares how the operation of the payments system was sustained in 1989 in the aftermath of the Loma Prieta quake. John Munro’s paper represents an extended reconsideration of M. M. Postan’s thesis about the impact of the Great Plague on wage rates. Munro insists that we systematically separate out the effect of real shocks which may have affected relative prices from monetary factors which influenced the general price level. His second major concern is to document and consider the consequences of nominal wage rigidity for real wages and living standards in the face of prolonged periods of deflation or inflation. If Munro explores the consequences of nominal wage stickiness in medieval England and the Low Countries, Andrew Seltzer examines wage flexibility half a millennium later in Australia in 1931. He uses personnel records from the Union bank to conclude that real wages were inflexible in the face of a governmentally mandated 10% wage cut, the consequence of an implicit contract between the bank and its employees. In our third paper focusing on the early 1930s, David Greasley and Jakob Madsen explore econometrically the debate about what caused the drop in
4259 Prels 4/12/02 9:41 am Page xi
Introduction
1 2 3 4 5 6 7 8 9 10111 11 12 13 14 15 16 17 18 19 20111 21 22 23 24 25 26 27 28 29 30111 31 32 33 34 35 36 37 38 39 40111
xi
consumption spending in the United States. They discount the influence of rising debt levels, the penalties for defaulting on installment contracts emphasized by Martha Olney, or changes in the supply of credit. Their conclusion instead is that income uncertainty underlay most of the spending declines between 1930 and 1932. Gary Richardson’s reconsideration of Deirdre McCloskey’s analysis of scattering in English open fields emphasizes that markets were much thicker and more fully developed than McCloskey assumed. He also finds evidence of many other mechanisms besides scattering for mitigating risk for all members of a village. Moreover, he finds little evidence of any efficiency losses from the open field system up until the eve of the Enclosures. Richardson’s analysis will be an important reference point for researchers in this area in the future, just as McCloskey’s seminal article – also published in Research in Economic History – has been for more than a quarter of a century. It’s a pleasure in this volume to publish work of scholars at all stages of their careers. We have contributions ranging from those of recently minted Ph.D.s who have excerpted and developed some of their thesis material for this journal, to those of distinguished senior scholars. Each of these articles is written with care, polish, and often passion. Academic disciplines flourish – and economic history is no exception – when scholars immerse themselves in their subjects and combine this with commitments to logic and evidence, detail, and clarity of exposition. The consequences are the fascinating papers evident here. We look forward to continuing to publish innovative, well written and carefully considered contributions to economic history, providing a niche which complements outlets such as the Journal of Economic History, Explorations in Economic History, and the Economic History Review. Potential contributors are urged to contact the editor for information on submission requirements. Alexander J. Field Series Editor
xi
4259 Ch01 4/12/02 9:42 am Page 1
1 2 3 4 5 6 7 8
MADE IN GERMANY: THE GERMAN CURRENCY CRISIS OF JULY 1931
9 10 11 12
Thomas Ferguson and Peter Temin
13 14 15 16
ABSTRACT
17 18 19 20 21 22 23 24 25
The Great Depression reached a turning point in the currency crises of 1931, and the German banking and currency crisis was a critical event whose causes are still debated. We demonstrate in this paper that the crisis was primarily domestic in origin; that it was a currency crisis rather than a banking crisis; and that the failure was more political than economic. German banks failed in 1931, but the problem was not primarily with them. Instead, the crisis was a failure of political will in a time of turmoil that induced a currency crisis.
26 27
I. INTRODUCTION
28 29 30 31 32 33 34 35
The currency crises of 1931 have emerged as pivotal events in the propagation of the Great Depression. It was during these crises that central banks and governments had to choose between domestic economic stability and maintenance of the gold standard. With the prominent exception of Great Britain, most industrial countries opted for the latter, dooming the world to the horrors of depression. To understand the Great Depression, one needs to understand the currency crises of 1931.
36 37 38 39 40
Research in Economic History, Volume 21, pages 1–53. Copyright © 2003 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623-0993-8
1
4259 Ch01 4/12/02 9:42 am Page 2
2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
THOMAS FERGUSON AND PETER TEMIN
The German crisis of July 1931 in particular has been a topic of study. Given Germany’s location, size, and subsequent ill-fated history, historians have attempted to disentangle the many factors inducing the German crisis. It was a twin crisis in the sense that there was both a currency and a banking crisis, rather like the Asian crises of 1997. The causes of this twin crisis have been found in many places by many authors – in domestic and in international events, in the currency market and the banking system, and in economic and political malfunctions. Historians have differed widely on the emphasis given to these various dimensions.1 For example, Kindleberger and Eichengreen emphasized the international aspect of the German crisis (Kindleberger, 1986; Eichengreen, 1992). James and Balderston, in contrast, stressed the domestic nature of the crisis (James, 1986; Balderston, 1994). We assert in this paper that the German crisis of 1931 was primarily domestic in origin, that it was a currency rather than a banking crisis, and that the failure was more political than economic. A currency crisis is national, affected by government policies and actions; a banking crisis is the result of private actions affected by fears for the safety of bank deposits. We argue that the German government found itself in an impossible position as the recession of the late 1920s deepened and its budget went into deficit. The Weimar government lacked the political will to deal with the impending crisis, substituting rash statements about customs unions and reparations for serious budgetary action. There is no evidence of international contagion from the Austrian crisis two months earlier; there is little evidence of structural problems among German banks or fear for their stability before the statements of the Weimar government induced fears for the value of the mark. German banks failed in 1931, but only because the Weimar government put them under unbearable strain. We document this assertion with monthly bank data and weekly Reichsbank data. We set the stage by first reviewing the literature on the German crisis of 1931 and showing that economic conditions in Weimar Germany during the first two quarters of 1931 were bad, but not getting worse. We examine banking data in that context to search for evidence of a banking crisis. We then review in detail the budgetary problems facing the German government in early 1931, showing how the government’s inability to resolve this problem amidst rising international tensions brought on the crisis. A conclusion places the German crisis into the narrative of the Great Depression as a whole.
36 37
II. EXISTING LITERATURE
38 39 40
Historians have formulated narratives of 1931 that explicitly and implicitly take positions on the cause of the crisis. The initial German response to the crisis
4259 Ch01 4/12/02 9:42 am Page 3
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
3
was to blame foreigners for withdrawing their deposits in Germany and selling marks.2 In this story, the withdrawals were triggered by changes in financial fortunes on Wall Street or by contagion from Austria. They were not occasioned by events within Germany. This story was international because it placed the movements of international “hot money” at the center of attention. It was a story about currencies rather than about banks, and it involved the German government not at all. Kindleberger may be taken as the start of the modern English-language literature on the European depression. He stated, “At the end of May 1931, Austrian financial difficulties ramified widely and led to runs on the banks of Hungary, Czechoslovakia, Romania, Poland, and Germany” (1986, p. 148). Kinndleberger’s subsequent discussion focuses on the consequent problems of German banks and makes no reference to German budgetary problems. In this context of bank fragility, the Nordwolle failed in mid-June, setting off the final stages of the crisis (Kindleberger, 1986, pp. 148–149). Kindleberger thus saw the German crisis as international in origin and mostly concentrated in banks. Eichengreen extended this internationalist tradition. He also argued that the Austrian crisis affected Germany through its effect on confidence, even though German banks were not heavily invested in Austria. Confidence in Germany collapsed because foreigners could not distinguish between Germany and Austria and because Brüning took the occasion of the Austrian default to insist on a reduction in reparations. Eichengreen did not mention the budgetary problems of the German government. Instead, he asserted that “the failure of international cooperation was the key to Germany’s defection from the gold standard system” (Eichengreen, 1992, p. 276). Eichengreen agreed with Kindleberger that the German crisis was international and carried the argument further by focusing on currency problems and failures of omission by governments outside Germany. James presented quite a different story with much more attention to internal German problems. He represented the crisis as a run on German banks that contained “structural weaknesses” and were “fundamentally unsound” (1986, pp. 294–295). The main cause of instability, in other words, was internal to Germany. The small push that toppled this unsound structure, however, was the withdrawal of foreign deposits, primarily American. International aspects were important in this story, but secondary if the crisis was the result of unsound German banking. James also argued that the Germans had brought the international crisis on themselves by tolerating capital flight in 1930–1931 and by mismanaging their debt. James only introduced the debt problem into his narrative of the crisis at the last moment, arguing that the budget crisis surfaced on June 9 when revenue figures for April and May became available, that is, 3
4259 Ch01 4/12/02 9:42 am Page 4
4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
at a late stage of the crisis (James, 1986, p. 306). In contrast to Kindleberger’s and Eichengreen’s story of international contagion, James’ story concerned German banks in a context of government financial problems. James recently articulated a view much closer to Kindleberger’s. In a book about the Deutsche Bank, he stated, “The Creditanstalt[’s] . . . failure precipitated a general banking and financial crisis in Austria that then spread to the whole of central Europe and eventually brought down German banks as well.”3 This view is consistent with his earlier emphasis on banks, but it places a much greater emphasis on international contagion than his earlier account. Balderston attempted to disentangle the banking and currency crises even as he asserted both were present. He argued that the proximate cause of the crisis was a general run on bank deposits due to conflicts over reparations and fiscal crises. The ultimate cause of the banking collapse was the inability of the German banks to replenish their reserves either from foreign loans or the Reichsbank. Balderston argued that the inability of German banks to borrow abroad was the root cause of the crisis, for only in that case did they go to the Reichsbank. Like James, Balderston focused on the problems of German banks, but like Eichengreen, he attributed the crisis to a failure of the international economy (Balderston, 1994). Balderston elsewhere presented a detailed account of the budgetary problems of the Weimar government in which he asserted that “the connection between the state’s cash crisis of June 1931 and the banking crisis which had been developing since May is too obvious to need elaboration” (1993, p. 312). But he treated the budget question and the currency crisis quite separately in his book, and he did not emphasize connections between them either there or in his subsequent paper. Balderston accordingly ends up with two rather disconnected stories of the crisis. His main story is about international aspects of German banking. But in the background runs a quite different story about the political problems of the German government. The two stories may be compatible, but they remain quite distinct. Consumers of these histories have two choices. They simply can conclude that all the above factors were important and take away a story in which all these factors – domestic and foreign, banking and governmental, economic and political – figure as separate causes. Alternatively, and preferably, they can try to disentangle this complex knot of explanations and search for a unified story. This story would not ignore any of the factors noted above, but it would state clearly where the crisis originated and specify the channels by which it spread to become a national and international tragedy. It is the purpose of this essay to provide such an account.
4259 Ch01 4/12/02 9:42 am Page 5
German Currency Crisis 1
5
III. ECONOMIC CONDITIONS IN EARLY 1931
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
The banking crisis of June and July 1931 was integral to another, even greater catastrophe – the collapse of the Weimar Republic and the Nazi seizure of power. The aura of impending doom that clings to all accounts of the period makes it easy to perceive disaster lurking everywhere, and creates a false impression of almost mythical inevitability. Accordingly, the first step in developing a unified account of the crisis is to acknowledge the old adage that history proceeds not in straight lines, but in spirals, by distinguishing as carefully as possible among the stages of the slide to the abyss. A crucial point is that at the beginning of 1931, the German economy was down, but not yet completely out. Not that the economic data, summarized in Table 1, were anything but ghastly: Industrial production had fallen to three quarters of its 1928 level. Unemployment had risen to over 30% of the unionized labor force, and the prevalence of part-time work had risen from its “normal” level of under 10% to about 20%.4 The special fund that financed the Weimar regime’s proudest achievement – the 1927 law establishing a system of unemployment compensation – teetered on the edge of bankruptcy as payouts ran far ahead of taxes paid into it. As the central government reacted by tightening eligibility, raising taxes, and reducing payments, the armies of the unemployed grew increasingly desperate. Many were recruited into the swelling ranks of paramilitary groups run by parties and groups on the right; others joined groups run by parties of the left, notably the German Communist Party (KPD). Those who exhausted their benefits or who could not qualify for any, could sometimes obtain assistance from the patchwork of local welfare programs run mostly by cities or town, but these programs were rapidly overwhelmed by the crush of applicants and financial exigency. With many cities and states (Länder) in desperate financial straits and clamoring for bailouts from the central government, community cohesion was severely strained.5 But the economic situation, while desperate, was more complicated than has often been recognized. The contraction appeared to be ending in early 1931 as economic conditions improved slightly. It is necessary to sketch briefly the political intrigues of 1930 in order to appreciate how this apparent halt to the economic decline affected the German government’s budgetary deliberations. In the Spring of 1930, only days after narrowly ratifying the Young Plan revision of the Dawes Plan, the Reichstag deadlocked over the budget. Advised by General Kurt von Schleicher, President von Hindenburg refused to allow the government headed by Chancellor Hermann Müller of the German Social Democratic Party (SPD) to make use of the President’s power to enact measures 5
11
13
14
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Source: Wirtschafts und Statistik, 1930–1931.
9
15.4 16.1 16.9 19.2 19.5 18.9 18.1 17.4 17.7
10
23.6 26 31.7 34.2 34.5 33.6 31.8 29 29.7
8
77.1 75.6 72.2 67.8 69.2 73.5 76.3 73.9 74.4
843 829 850 1085 1065 1240 972 956 1034
10,191 10,272 10,400 10,767 10,946 11,064 11,165 11,225 11,074
6
October1930 November 1930 December 1930 January 1931 February 1931 March 1931 April 1931 May 1931 June 1931
17 Savings Bank Deposits (Million RM)
5
Month
7
Bankruptcies (Monthly Total)
4
Partially Unemployed (% Trade Unionists – See Note 4)
3
Unemployment (% Trade Unionists – See Note 4)
2
Industrial Production (1928 = 100)
16
German Economic Data.
15
Table 1.
4259 Ch01 4/12/02 9:42 am Page 6
THOMAS FERGUSON AND PETER TEMIN
1
6
12
4259 Ch01 4/12/02 9:42 am Page 7
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
7
by decree under Article 48 of the Weimar Constitution. When the government resigned, the President, a former general and hero of the battle of Tannenberg, who had personally helped mint the “Stab in the back” legend after World War I, offered the Chancellorship, and the use of Article 48, to Heinrich Brüning of the Catholic Zentrum party.6 Brüning, who, in Eberhard Kolb’s words, “from start to finish regarded himself as holding office solely by virtue of the President’s authority and in order to fulfill the mission assigned to him,” immediately ran into heavy weather in the Reichstag, which could still overturn Hindenburg’s emergency decrees (Notverordungen) by majority vote.7 The major stumbling block was the same issue as before: production and trade were falling as expenditures for relief and unemployment compensation soared, unbalancing the budget and threatening further rapid increases in government debt at all levels (federal, Länder, municipalities). As shown in Table 2, Germany’s total government debt rose from 15 billion RM at the end of March 1928 to 24 billion RM in 1931.8 The federal government accounted for about half of the consolidated total, indicating that all levels of government were running deficits as the economy declined.9 While total foreign debt had risen, it did not comprise the bulk of government debt. In addition, relatively little of the official foreign debt was short-term. Most government debt was held domestically, and the volume of short-term debt rose with the total. (The sum of foreign and domestic debt falls short of the total by the amount of “old debt.” As can be verified by subtraction in Table 2, the volume of “old debt” did not change.) Pressed to get this mountain of debt under control, Brüning enacted by Notverordung a budget calling for savage spending cuts, lower taxes, cuts in the pay of government officials, and smaller transfers to the Länder and municipalities. The move outraged many on both the left and the right for different reasons. After the big business oriented German People’s Party forced the resignation of the Finance Minister and the conservative German National
30 31
Table 2.
German Government Debt (Million RM).
32 33 34 35 36 37 38 39 40
All governments Reich only Foreign debt Foreign Sh Tm Domestic debt Domestic Sh Tm
3/31/28
3/31/29
3/31/30
3/31/31
14,599 7,139 2,157
18,159 8,229 2,294 185 9,263 2,569
21,319 9,630 2,595 532 12,415 4,070
24,022 11,342 4,799 830 13,181 3,746
5,547
Source: Länderrat, 1949.
7
4259 Ch01 4/12/02 9:42 am Page 8
8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
People’s Party (DNVP) committed binary fission, a majority of the Reichstag exercised its constitutional right to reject the decree. Hindenburg and Brüning reacted with a fateful step: They dissolved the Reichstag, called new elections, and re-enacted the Notverordung with minor changes.10 A disaster for the government ensued. In the September elections, voter turnout surged and the Nazis dramatically emerged as the second largest single party in the Reichstag, while the Communists made substantial gains (Winkler, 1993, pp. 388ff.). Parties supporting the governing coalition now controlled only about a third of the Reichstag, with another third dominated by parties fundamentally opposed to the existing political system.11 Not surprisingly, investors stampeded out of Mark assets. In a single month, 800 million Reichsmarks worth of foreign capital left the country, while German securities fell sharply in foreign markets. The Reichsbank was forced to raise its rediscount rate by a full point.12 It is here that the temptation to write the history of the 1931 crisis in straight line terms has been most meretricious. Many accounts treat the crisis of June and July as a seamless projection of the late fall, 1930 panic. In some versions, the capital outflow never really abated. According to others, the 1930 crisis closed world capital markets to German enterprises and, especially, the German government, making it impossible for the government to service either its foreign or domestic debts. A recent view suggests that the overhang of Young Plan loans, which had priority for repayment in a crisis, made it impossible to float new ones.13 What actually happened is that the government responded by redoubling its resolve to continue down the path of austerity so resoundingly repudiated by the electorate. Amid a flurry of sympathetic news stories in papers from Berlin to New York, Hindenburg immediately tapped Brüning to form a new government. Solemnly promising to make fiscal reform and deficit reduction his top priority, Brüning reappointed the entire cabinet. On the basis of more Notverordungen and additional measures mandating a sinking fund for debt retirement, steep rises in workers’ contributions to the unemployment fund, higher taxes (including new levies on mineral water and individual citizens), and slashes in government spending, civil service pay, and inter-governmental transfers, the government floated a $125 million loan through a syndicate headed by Lee, Higginson.14 The spectacle of the government bowing to foreign creditors (whose connection to the austerity package was too obvious to escape acrid public comment at home) reduced the government’s popularity still more. But fears of what would happen if Brüning’s government collapsed forced the shell-shocked SPD into an agonizing reappraisal. Staring into the abyss, the party reluctantly
4259 Ch01 4/12/02 9:42 am Page 9
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
9
concluded that while it could not openly support the Chancellor, it would have to “tolerate” his government as the lesser evil. When the Social Democrats flirted with the idea of voting down some of the budget measures, the Chancellor brought them into line by threatening a Zentrum pullout of the SPD-led government in Prussia. In the murderous climate of the times, the SPD simply could not afford to lose control of the police force that controlled the largest of the German Länder.15 With this new, wider support, the Brüning government moved to reassure financial markets anxious about the domestic political scene and trial balloons the government had discreetly floated about possibilities for further concessions on reparations. Encouraged by the British government and the Bank of England (whose Governor, Montagu Norman, endorsed a plan to have the new Bank of International Settlements inject liquidity into central European financial markets), the Brüning government and the Reichsbank joined forces to strike back at former Reichsbank head Hjalmar Schacht and other critics who were campaigning for unilateral revision or repudiation of the Young Plan. Flatly ruling out action in the near future, Brüning instead preached that further reparations revisions would have to wait until Germany demonstrated to the rest of the world that the limits of her ability to pay had truly been reached by reining in the budget and reforming the fiscal system.16 Both governments and financial markets swiftly perceived that “while Dr. Brüning certainly continues to cherish hopes that, by careful internal and external preparation, he will eventually secure, by mutual consent, a further reduction of Germany’s reparation liabilities . . . [he] has become thoroughly convinced that the disadvantages which would arise from an appeal to the safeguarding clauses in the Young plan far outweigh the partial and temporary relief which Germany would obtain therefrom, and is in no mood to shortsighted pressure in this direction.”17 The government’s show of resolve rallied support in and out of Germany. In speeches that made news even in New York (the financial center with the most money of all invested in Germany), Brüning explained “like a patient and firm schoolmaster” that after cutting spending, the government would seek to consolidate its huge short term floating debt into longer term obligations (New York Times, February 6, 1931, p. 9).
35 36 37 38 39 40
In public speeches both before and after the Reichstag met on the 3rd of February the Chancellor showed courageous frankness and his increasing confidence infected the leaders of the Government parties . . .. The large majorities obtained by Dr. Brüning in the opening days of the present session of the Reichstag again impressed public opinion with his qualities of leadership, and showed that the other political parties had at last realized the necessity of standing up to the extremists.18
9
4259 Ch01 4/12/02 9:42 am Page 10
10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
International support for Brüning grew as his government made good on its tough talk. In January, the International Institute of Finance, a think tank with strong ties to financiers interested in European debt, provided timely reassurance with a widely heralded survey concluding that foreign investors need not fear that Germany might be forced to repudiate private debts in the new year. The British Foreign Office started exploring ways “His Majesty’s Government” might “give [Brüning’s government] such support and encouragement as they properly can, in order to fortify its position.” BIS officials delivered speeches calling for an increase in international lending and promising to lead the way in central Europe with the bank’s own money.19 With confidence reviving, official statistics on capital outflows and banking positions improved markedly. Studying the figures, the Economist considered that they gave “every sign of returning calm” and “the cessation of the withdrawals of money, especially foreign withdrawals.” Private bank reports mostly echoed the public media, if initially somewhat more guardedly. In a mid-January report that was certainly never intended for wider publication, Morgan et Cie., the Paris affiliate of the most influential bank in the world, reported to its New York parent that “German bankers report money beginning to return to Germany and they anticipate some weakening of rates.”20 Somewhat later the Reichsbank cabled essentially the same message to the New York Fed.21 Other favorable developments bolstered the growing confidence. First and foremost was the fact that the balance of payments was firmly in the black. Germany had been importing capital steadily since the stabilization of the Reichsmark. It had in fact borrowed considerably more than its reparations obligations, leading to what Schuker (1988) called American reparations to Germany. While the net capital inflow decreased in 1930, the decline in capital imports was offset by a fall in imports of goods and services. Relevant data are set out in Table 3. Exports reached their peak in 1929 and fell thereafter, but imports fell faster, leading to a rise in net exports. Capital imports also fell in 1929 and 1930. These two movements largely offset each other, as of course they must to make the balance of payments add up. (Other, smaller items in the balance of payments did not change as rapidly as these categories.) Net income, the resultant of all these forces, was remarkably stable. The need to pay reparations meant that Germany’s balance of payments was not in equilibrium when net imports equaled the capital inflow, but rather it was in equilibrium when the capital inflow exceeded net imports by the amount of reparations payments, which amounted to between one to two billion RM, depending on the time period one reckoned over. This represented a shift of the location of equilibrium and a burden on Germany, but it did not change the short run dynamics of the balance of payments. It was German reluctance
4259 Ch01 4/12/02 9:42 am Page 11
German Currency Crisis 1
Table 3.
11
German Balance of Payments, 1928–1931 (Million RM).
2 1928
1929
1930
1931
13,599 13,912 313 6,958 2,852 4,106 0 931 931 2,862 1,999
14,344 13,676 668 4,459 1,991 2,468 510 345 165 3,301 2,501
12,713 10,617 2,096 3,678 3,195 483 192 72 120 2,699 1,699
10,169 6,955 3,214 3,817 6,494 2,677 1,653 0 1,653 2,190 990
3 4 5 6 7 8 9 10 11 12 13 14
Exports Imports Net exports Capital inflow Capital outflow Net inflow Gold exports Gold imports Net gold exports Net income Reparations
Source: Statistisches Jahrbuch, 1933, pp. 498–499, as quoted in Schmidt, 1934.
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
to pay reparations in the longer run that made them a critical political issue.22 In early 1931, amidst the improving payments situations, the economics of the transfer issue did not seem insuperable. The consensus view was that “there will be no difficulty in making transfer” of the payments due under the Young Plan.23 Other economic statistics trickled in that fanned this optimism. Widely reported indices of German industrial production rose slightly, as can be seen in Table 1. Unemployment began to trend very slightly down. Bankruptcies – even now esteemed as one of the few clear external indicators of possible bank lending difficulties – started falling in February, rose slightly in March, and then fell sharply in April and May (Table 1). Indeed, almost every indicator that subsequent analysts have suggested as heralding banking or currency crises improved modestly in the early months of 1931. With interest rates – even long rates – falling, the German stock market rose for four months, while the bond market – including, notably, the internationally traded Young Plan bonds, surely a prime index of global sentiment toward Germany’s repayment possibilities – rallied sharply, as shown in Table 4.24 In an age that still venerated gold as the ultimate test of sustainability, Reichsbank holdings of gold and foreign exchange mounted slowly, but steadily. From 2216 million RM. at the beginning of 1931, Reichsbank gold reserves rose to 2323 million RM by the end of March. In January, Reichsbank holdings of foreign exchange dropped from 400 million to 200 million RM. As the Brüning government redeemed its budgetary promises, the outflow slowed and then reversed (see Table 8). By the third week of March – a fateful date, as 11
11
12
13
14
17
18
19
20
21
22
23
24
25
26
27
29
30
31
32
33
36
37
38
39
40 85.58 85.85 85.78 86.64 87.46 87.32 86.28
– 80.51 80.61 81.58 82.47 81.7 78.44
70.5 72.75 78.75 80 80.63 70 75
Young Bonds
1
4
28
34
35
Sources: Stocks, Overnight Money, Gold Bonds (monthly averages): Deutsches Institut für Wirtschaftsforshung, 1932, p. 9. Bank Stocks, Public Bonds (monthly averages): Reichsamt, 1932b, pp. 354–355. Young Bonds (value at end of week nearest end of month); See Note 93 on June entry): Commercial and Financial Chronicle, 1931.
5.66 5.1 5.78 5.31 5.88 5.64 7.07
9
– 110.52 111.1 112.65 112.61 106.09 100.58
8
87.3 81.8 85.6 91.1 92.4 83 75.9
7
6% Public Bonds
12
December 1930 January 1931 February 1931 March 1931 April 1931 May 1931 June 1931
10
6% Gold Bonds (Gold-pfandbriefe)
6
Month
5
Overnight Money (Tagesgeld) Percent
3
Bank Stocks (Kredit-banken) 1924–1926 = 100
2
All Stocks, 1924–1926 = 100
16
German Financial Data.
15
Table 4.
4259 Ch01 4/12/02 9:42 am Page 12
THOMAS FERGUSON AND PETER TEMIN
4259 Ch01 4/12/02 9:42 am Page 13
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
13
will shortly become evident – the rise in the Reichsbank’s total holdings of gold and foreign exchange appeared to bear out the projections of the The Economist cited earlier. Inevitably, some optimistic commentators hailed these modest upticks as evidence that the invisible hand was at last beckoning the German economy on. It is not necessary to share this touching faith – or its corollary that only subsequent political disaster aborted recovery – to acknowledge that in the early months of 1931 “authoritative banking opinion” held that “the Reich will be able to borrow sufficient funds to meet her domestic and international obligations” for the coming year. The international bankers who have had occasion recently to investigate the conditions of Germany are optimistic regarding the future. Complex problems confront the German government, but in Dr. Heinrich Brüning they see a quiet, forceful leader, capable of guiding the nation through a difficult period. His desire to permit nothing of a domestic nature to influence the calm, reasoned course of Germany’s foreign relations is accepted by the bankers as a fairly sure guarantee of a new period of stability (New York Times, Feb. 8, 1931, p. 8).
In February, through a syndicate led once again by Lee, Higginson, the government successfully returned to international money markets, borrowing RM 120 million for eighteen months to cover the deficit Brüning was pledged to reduce.25 As the Nazis and the strongly nationalistic DNVP walked out of the Reichstag in protest, giving Brüning a comfortable working majority, he momentarily acquired a stature almost larger than life. Leading international dailies echoed the views, quoted earlier, of the British ambassador, affirming that the Chancellor was “stronger than ever,” while their business pages gushed about “the particularly welcome improvement in the Berlin money market due to the continuing inflow of foreign funds, largely in the form of three month credits . . .. Bankers express the view that the ‘confidence crisis’ is ended” and that “further foreign credits will come in and correct the wide disparity between rates at home and international capital markets” (New York Times, March 23, 1931, p. 33). The revival of confidence reached its zenith in March. Highlighting his government’s success in reducing the level of short term debt, Brüning succeeded in passing his budget through normal parliamentary processes (i.e. without recourse to Notverordungen). Then he persuaded the Reichstag to recess until October. [This] probably secures the country against an internal political crisis in the next few months. The attitude of the foreign banks has correspondingly improved and money has flowed into Germany from Holland, Switzerland, France, and Belgium. Berlin banks accepted the funds at reduced rates, but in general only to maturities of three months, while they paid
13
4259 Ch01 4/12/02 9:42 am Page 14
14 1 2 3 4
THOMAS FERGUSON AND PETER TEMIN one-month credits. This policy is due to the banks’ conviction that they will soon get foreign money cheaper.26
One telegram in particular is uniquely authoritative about what “markets” were thinking about Germany at that time:
5 6 7 8 9 10 11 12
Though it has not yet reached the stage where any large foreign offerings are immediately practicable, there has already been a considerable improvement in the market position of many foreign issues and the German Government 51/2s as you will have noticed have improved from about 73 around the 1st of February to about 82 at present quotations. This reflects the market’s greater confidence in the German situation and in the prospects for political security in Europe as well as the assurances given by Dr. Luther and prominent German officials to the effect that Germany has no intention of stopping payments on her obligations to private investors, including specifically the Dawes and Young loans.27
13
A few days later, one of the authors of this rosy scenario went a step further:
14
Some years ago, I made a bet that French credit would rise to a 41/2% basis on our market. That bet I have collected. It emboldened me to make another which was that the German 51/2s would sell at par by January 1, 1935. I feel confident that if I live that I shall collect this bet also.28
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
The most dramatic development in regard to Germany’s potential creditworthiness during this period has been almost entirely forgotten. In February came sudden news that the French were reconsidering their previously hostile attitudes toward possible long term loans to Germany. Save for the Young Plan loan, French banks had not participated in any major credit to Germany since 1929. Deeply alarmed by results of the September election, they had declined any part of the earlier Lee, Higginson credit. Paris newspaper accounts at that time intimated that France, then the second largest creditor nation in the world thanks to the enormous inflow of gold that followed the Poincaré government’s undervaluation of the Franc, possessed the financial clout to bring down the German economy by calling in loans.29 In December, however, the conservative, nationalist government headed by Andre Tardieu collapsed. After a brief caretaker regime, a new government headed by Pierre Laval took over with Aristide Briand, the long time champion of Franco-German cooperation, serving as Foreign Minister. Not for the last time in French history, a Laval-led government tilted in the direction of its bigger neighbor. Briand, who had his eye on the election for the Presidency of France scheduled for May, 1931, had for some time been pushing for a change in policy to head off the rise of a still more nationalist government in Germany by easing tensions and reviving the economies of both countries. Laval was also working with important Belgian bankers and diplomats, who were anxious over the steady deterioration of Franco-German relations. In both Berlin and Paris, Lee, Higginson also actively promoted rapprochement, not least by taking
4259 Ch01 4/12/02 9:42 am Page 15
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
15
on a senior French banker with glittering policy credentials, who was close to many top French policy makers.30 Gradually a disbelieving world awoke to the fact that something remarkable was happening. First came hints that Lee, Higginson was once again considering a large loan to Germany. A few days later came the sensational announcement that at the request of the French government, the giant Banque de Paris et des Pays Bas would participate in the credit. This news “made a profound impression on the Paris bourse, where the move was widely interpreted as the initiation of a new foreign credit policy.”31 At the General Assembly of the shareholders of the Banque de France, Governor Moret “pointed to the possibility of more French lending abroad and urged the reduction of restrictions in the bond market.” In a speech in Paris on February 12, BIS President Gates McGarrah warmly endorsed Moret’s views, triggering a lively discussion of how recycling stored up French funds could help restore international financial equilibrium.32 The sudden realization that the Paris market might be far more lucrative than anyone could have imagined only a few months before momentarily threatened to disequilibrate the market for German goverment debt – but in a direction very different from that emphasized by recent German historians. Drawing mainly on official records and papers left by senior civil servants, James and, especially, Bachmann, paint decidedly pessimistic pictures of that market in this period.33 Since only a short time later demand for German government obligations did indeed collapse, annual statistics for total loans inevitably appear to validate their pessimism. But the archives of private bankers reveal a much more complex and interesting situation, brought on by widespread interest among investment bankers in German debt. Lee, Higginson, the lead bank in most of the international syndicates that brought out German government bonds at this time, was acutely aware that several rivals were after the business. It also understood that its position in Germany depended crucially on its ties to Mendelssohn & Co., the legendary Berlin private banking house. This firm was close to many prominent officials in the German government, including Reichsbank President Hans Luther and Hans Schäffer, who held the key post of State Secretary in the Ministry of Finance.34 As prospects for new German loans brightened, Mendelssohn became restive and started pressing Lee, Higginson for a better deal. It first sought preferences in northern markets where it had historically been strong. This boomeranged when a major Swedish bank made its participation in future syndicates conditional on the leadership of Lee, Higginson. In mid-February, however, with French banks flocking to take up parts of the new German credit, the 15
4259 Ch01 4/12/02 9:42 am Page 16
16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
THOMAS FERGUSON AND PETER TEMIN
German bank shocked its American partner by declaring “positively that the Paris market must be reserved for Mendelssohn and Company, Amsterdam” because of its “close connections with leading French institutions, particularly Banque de Paris et des Pays Bas.”35 At the time, Lee, Higginson was campaigning to persuade the German government to designate it as the government’s exclusive manager for future internationally syndicated loans. Still, the American bank, which had its own Paris affiliate and was trying to build up its business, flatly refused its German ally’s demands. In the end, “certain domestic factors, principally the opposition of Schacht” dashed the American house’s hopes of an exclusive franchise.36 Opposition from some key officials in the Reichsbank also blocked other concessions Lee, Higginson proposed to placate Bank Mendelssohn until the whole struggle over expected future profits from French loans to Germany became moot. But these rivalries over the German government’s debt offerings strikingly indicate just how little economic conditions in the early months of 1931 hinted at the crisis that was just around the corner. If conditions were not good in these harsh economic times, there was little anticipation that they were about to get a lot worse. Indeed, expectations appeared to argue the opposite, that conditions were on the mend. Yet in the second quarter of 1931, something went wrong. The question is whether the change was in the stability of banks or of the Reichsmark, in Germany or without, in politics or economics.
22 23
IV. THE BANKING CRISIS
24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
Once the crisis began, panic spread and everything fell apart. But describing the crisis does not help us discriminate between its putative causes. We must examine the process that led up to the crisis. In this section we present banking data to distinguish the effects of the budgetary crisis from two other antecedents of the crisis whose effects also have been touted in the literature: the Austrian currency crisis in May and the weakness of German banks. The tests are rather different. To see if the Austrian crisis mattered, we examine the German reaction to the Austrian crisis. Did the Austrian crisis set off a sequence of events that led to the German crisis in July? To see if banks were unstable, we examine various indicators of bank performance in the months before the German crisis to see if there were signs of distress. The source material is almost the same for the two tests. The difference is that the first set looks specifically at timing, while the second examines banking indicators in a longer-run perspective. There is an extensive literature about German banks, as noted above, but there is not a clear idea what to look for. At the risk of narrowing our focus
4259 Ch01 4/12/02 9:42 am Page 17
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
17
excessively, we make explicit the implicit model of other scholars. When we talk about a banking panic, we refer to an American-style banking panic.37 In this kind of panic, depositors become fearful for the safety of their deposits. They withdraw their deposits, starting with the most liquid since they are in a hurry to get their funds into a more secure asset – cash as opposed to bank deposits. Banks operating with fractional reserves have to draw down their reserves in order to pay out deposits. Their reserve ratio falls. They call in loans to rebuild their reserves. Being short of reserves, they exert pressure on borrowers to pay up as soon as possible. In order to pay off these loans, borrowers need to withdraw their deposits from the banks where they keep money. The bank withdrawals that may have started with a single bank or a small number of banks then spread to all banks as the ones affected initially call in loans that force other banks to pay out deposits. The banks draw down their reserves as they pay off their deposits, forcing them to call in their loans. Soon all banks are calling in loans. Everyone is scrambling for liquidity, and there is not enough liquidity to go around. There are two important steps in such a banking panic. First, scared depositors draw out their deposits. Since the spread of a banking panic is caused by the attempts of banks to rebuild reserves quickly, a marker of a panic is that people withdraw money from sight accounts. If they withdraw term deposits, they give the bank time to reconstruct its reserves without imposing as much pressure on other banks as when they need immediate help. Second, banks call in loans in order to rebuild their reserves. If they have some other way of getting reserves – say, from a central bank – they do not transmit pressure to other banks and initiate and intensify a general banking panic. We examine German banking data to see if these two conditions were present. The following tables contain relevant banking data from 1931.38 Table 5 reports deposits by bank class. The six Berlin Grossbanken are listed first. Kreditbanken are listed next. Finally, the Kreditbanken other than the Grossbanken are obtained by subtraction. Deposits are divided between those accessible within a week, labeled demand deposits (DD), and those accessible in eight days to three months, labeled time deposits (TD).39 Table 6 contains data on the largest assets of the Grossbanken.40 Table 7 reports time deposits for individual Berlin great banks.41 The first thing to note is that demand deposits did not fall at all in the crisis. If there had been a banking crisis, one would have expected these deposits to be withdrawn faster than all others. They were the easiest deposits to withdraw, and depositors should have converted them to cash or foreign exchange at once. Their stability suggests calm rather than panic. We do not mean to argue that 17
4259 Ch01 4/12/02 9:42 am Page 18
18 1
THOMAS FERGUSON AND PETER TEMIN
Table 5.
Bank Deposits by Access (Million RM) – by Bank Class.
2 3 4 5 6 7 8 9 10 11 12 13
All Grossbanken DD TD All Kreditbanken DD TD Kredit. – Gross. DD TD
28/2/31
31/3/31
30/4/31
30/5/31
30/6/31
31/7/31
3756 4627
3819 4666
3657 4801
3626 4632
3626 3519
3891 2370
4528 5578
4562 5668
4390 5850
4382 5611
4327 4432
4749 3152
772 951
743 1002
733 1049
756 979
701 913
858 782
Source: Wirtschaftsdienst, 1931, various issues.
14 15 16
Table 6.
17 18 19 20 21 22 23 24 25 26 27
Schecks, Wechsel Remboursekredite Total Short-Term Loans against securities against physical assets Deposits at other banks
Assets (Million RM) of the Grossbanken. 28/2/31
31/3/31
30/4/31
30/5/31
30/6/31
31/7/31
2497 2006 5896
2530 1894 5890
2528 1828 5834
2547 1781 5734
1914 1748 5668
1280 1599 5484
1307 3195 946
1309 3214 856
1276 3219 981
1235 3183 857
1228 3071 686
1055 3014 546
Source: Same as Table 5.
28 29 30 31
Table 7.
Time Deposits (Million RM) – Individual Grossbanken.
32 33 34 35 36 37 38
De-Di Dresdner Danat Compri Reichskredit Berliner
39 40
Source: Same as Table 5.
28/2/31
31/3/31
30/4/31
30/5/31
30/6/31
31/7/31
1778 871 977 621 181 198
1789 874 989 639 176 199
1799 931 1038 650 187 197
1760 869 967 533 215 187
1444 631 647 518 133 145
1009 490 300 363 160 109
4259 Ch01 4/12/02 9:42 am Page 19
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
19
conditions in Germany were at all calm in June 1931; rather we argue that fears of that time were not directed toward large banks. The second thing to notice is that time deposits also stayed steady, although not so long. They were steady through the end of May. In fact, they were in the range that they had been in prior years. They ranged from 4,359 million RM in May 1929 to 5,682 in August 1930. Demand deposits also were in the range of the immediate prior years: 3,434 in May 1929 and 4,289 in September 1931 (Balderston, 1994). The stability of deposits through May indicates that the failure of the Creditanstalt did not lead to a panic of German banks. The third conclusion is more complex. Time deposits fell sharply in June 1931. As shown in Table 7, all Grossbanken lost time deposits. But the fall appears to have been concentrated in these six banks; the data in Table 5 show that time deposits in other credit banks did not fall nearly as much. If this fall is a sign of panic, it suggests a peculiar sort of panic. It extended to all the great banks, but not beyond. There is no sign of the contagion from bank to bank that is the hallmark of a general banking panic. Instead we find depositors shifting from longer to shorter loans, from less liquid to more liquid assets, as if they were preparing for a possible future problem. This does not suggest panic, but rather precautionary activity by depositors on a small enough scale not to threaten any bank’s stability. This movement does not indicate a loss of faith in banks; it suggests instead that some depositors feared for the value of the mark. This conjecture is supported by the data on bank assets in Table 6, reporting the assets of the six Grossbanken. As shown in Table 5, deposits of these banks fell by 1,000 million RM in June. Assets must have fallen by the same to balance their books. The table shows the largest categories of deposits; other smaller categories did not fall enough to matter. Loans to business firms did not fall, although there was a small decline of loans against physical assets, i.e. loans against the security of the firm as a whole. Checks and bills of exchange (Wechsel) fell by 600 million RM, accounting for almost two-thirds of the fall in deposits. Deposits at other banks fell by 200 million RM, and acceptances held by the banks (reported under liabilities) rose by the same amount. There is no sign of banks calling in loans, only of a decline of short-term financial assets as the volume of trade collapsed. The fall in deposits at other banks was small. It was also offset by a fall of roughly the same amount in the deposits from other banks included among the deposits in Table 5. The Grossbanken had very small net positions in deposits to and from other banks, and their exposure did not change. Even the fall in the gross position was only one-fifth as large as the fall in time deposits. This means that the fall in time deposits was not restricted to deposits from other 19
4259 Ch01 4/12/02 9:42 am Page 20
20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
banks, and changes in interbank funds did not create hardships for the Grossbanken. The rise in acceptances is mysterious. As noted, this is not a large change, but it contrasts with the fall in bills of exchange held and the decline in time deposits. British banks had substantial short-term loans outstanding to Germany; many of these were in the form of acceptances in order to earn a higher rate of interest than on deposits.42 As a result, any bank problem from the fall in time deposits in June was a smaller one than the deposit data suggest since the rise in acceptances offset part of the fall in time deposits. Nothing in these data suggests over-expansion or other excessive activity on the part of the Grossbanken. The pattern on the asset side does not look more like a bank panic than the pattern on the debit side. Fewer bills of exchange reflect changes in the international markets, whether of trade or of currency. Trade was declining as the Depression deepened, and traders began to hold their assets outside of Germany as they feared for the Mark. None of this suggests a banking panic. There is no contraction of loans by banks under attack which weaken other banks which in turn come under attack. There was no rush to withdraw the most liquid deposits and no inter-bank pressure to recoup reserves. Comparing these data to the pattern of banking crises shows that banks were not calling in loans that forced borrowers to withdraw deposits at other banks. The Grossbanken were not acting in a banking panic. The Danat bank experienced more withdrawals in June than some other great banks, but it lost proportionately slightly less than the government-owned Reichskreditgesellschaft (RKG), as shown in Table 7. Depositors do not seem to have singled it out as dangerous in June, despite its deep involvement with the Nordwolle failure. Depositors did draw more out of the Danat bank than from RKG in May and July, but it is hard to know what they were responding to. Nordwolle was not expected to fail in May, and deposits were guaranteed in July.43 For whatever reason, time deposits at the Danat bank had fallen the most of any great bank from the end of March to the end of July. Demand deposits at the Danat bank, however, had not fallen at all – more correctly, they fell slightly and recovered in July. These data support two conclusions. First, the collapse of the Creditanstalt did not affect German banks. There was no change in either the liability or the asset side of bank balance sheets in May. Indeed, there is no sign that the Austrian crisis had any effect on German banks at all – the Reichsbank was simply stating the truth when it privately advised the New York Fed right after the collapse that “our market is scarcely influenced by the reorganization as neither the Oesterreichische Credit-Anstalt für Handel und Gewerbe has large
4259 Ch01 4/12/02 9:42 am Page 21
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
21
credit balances here, nor [were there] considerable demands of our market on Vienna. Therefore money market calm.”44 Second, there is also no evidence of longer-run problems with the Grossbanken as a group. The data in Tables 5–7 suggest nothing so much as banking stability before June. The fall in time deposits in June may have been the result of depositors preparing for a future crisis by shortening the maturity of their deposits. But even this is not a clear signal. As noted already, this net movement was the result of offsetting changes, and there is no sign of panic in the steadiness of demand deposits. Due to the articulated structure of the German banking system, tests on these data are incomplete. American banks held deposits of both firms and individuals. If ordinary people feared for their bank balances, their actions would show up in the data on demand deposits. German banks were more specialized. The Grossbanken held deposits of business firms and prosperous individuals, while other types of banks held those of ordinary people. The data in Tables 5–7 therefore show that the business community was not frightened by bank behavior, but they do not indicate if people in the street were anxious. For that, we have to turn to data for other types of banks. Data for smaller savings banks, whose business was local and accounts were almost entirely in Reichsmarks, are revealing. All through the winter and spring of 1931, small savers were not fleeing their local banks. Indeed, at the end of May, three weeks after the announcement of the failure of the Creditanstalt that allegedly triggered the banking crisis, German savings banks actually held more deposits than they did at the beginning of the month. Only in June is there any sign of withdrawals, and the losses were not large. Roughly the same is true of the so-called “cooperative” banks.45 Finally we need to examine the deposit to currency ratio to make sure that the stability of bank deposits did not mask a shift of monetary holdings into cash caused by distrust of banks. Again, the parallel with the United States is not exact. Germans economized on cash through use of the Giro system, and the data for currency in Germany do not refer to the same part of the economy as the American system.46 While conventional accounts of the banking crisis do not focus on the types of banks that held deposits of ordinary people, the deposit-currency ratio may still be informative. It fell about five percent in April, but remained constant in every other month in the first half of 1931. There was no cumulative decline in the deposit-currency ratio as there was in the United States.47 Could the “banking weakness” story be saved by abandoning all reference to the Creditanstalt, switching models, and pointing away from American style bank panics to Japanese style slow motion asset deflations? While accounts of the crisis by Balderston, James, and others all take some kind of banking panic for granted, 21
4259 Ch01 4/12/02 9:42 am Page 22
22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
THOMAS FERGUSON AND PETER TEMIN
their emphasis on long building banking weaknesses does not fundamentally involve – save perhaps in the final stages – appeals to chain bankruptcy at all. As the contemporary example of Japan reminds us, banks can also collapse and cripple an economy through the slow decay of their assets and corollary inability to lend (Goldstein, Kaminsky & Reinhart, 2000, p. 20). The Nordwolle failure did bring down the Danatbank, and other large German banks, including the giant Dresdner, also failed. Still other large banks, such as the Deutsche Bank, appear to have had narrow escapes. We therefore cannot dismiss flippantly the notion that the German banking system was structurally fragile after three or four years of depression. Perhaps the large banks took excessive risks out of the conviction that they were too big for the Reichsbank to allow them to fail.48 It is not easy to tell if this story about deteriorating assets is true in the absence of archival evidence about bank portfolios. But the available evidence is less than overwhelming. We have seen that savers were not running on the banks as the crisis approached. They evidently did not sense collapse coming. Neither do bank stocks, one of the few external indicators of the onset of banking crises recommended by contemporary analysts of financial vulnerability, appear to provide support. While no comprehensive review of the stock market evidence is known to us, existing stock indices do not indicate unusual sell offs of bank stocks relative to other sectors (see Table 4). Since the banks were widely suspected of buying their own stocks to keep up their price, this evidence is suggestive, rather than conclusive, but the general rise in the stock market through April has to have helped banks, since they held large blocs of industrial shares.49 As late as May 9, 1931, The Economist reassured readers that German banks’ “gross profits have not declined very substantially; their decline has been counterbalanced by savings in expenses and taxation.”50 Another article in the same issue favorably compared the level of capital and reserves of the German banks to those of the British (The Economist, May 9, 1931, p. 8). Bankruptcy statistics – the other obvious external indicator of asset unsoundness – were also headed in the wrong direction (Table 1). By contrast, the devastation wrought by the currency collapse is clear. Without the run on the large banks’ foreign deposits induced by the currency collapse, the Danatbank and the rest of the banking system would have been far better armored against the Nordwolle shock. The Reichsbank would have had ample means to step in quickly, guarantee the deposits, and abort the final panic.
36 37
V. THE CURRENCY CRISIS
38 39 40
Why, therefore, were depositors worried in the late Spring of 1931, if they were not anxious about the soundness of German banks? Our answer is straightforward:
4259 Ch01 4/12/02 9:42 am Page 23
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
23
They were alarmed by the latest manifestation of the Weimar Republic’s long running fiscal crisis. This now threatened to go completely out of control and lead – for reasons of domestic politics – to a moratorium or repudiation of Germany’s foreign currency debts and to default or postponement of payments even on domestic debt. The first step down the fatal path came with the decision of Hindenburg and his advisers to push Brüning towards a more aggressive and nationalistic foreign policy. Soon after taking the reins, the Chancellor ostentatiously snubbed a proposal advanced by French Foreign Minister Briand to broaden the “Locarno System” of European security arrangements.51 Tensions heightened late in 1930, when the German government denounced the Draft Disarmament Convention for reaffirming the Versailles Treaty’s restrictions on German rearmament, and putting little pressure on the other major powers to fulfill their own pledges to disarm.52 More ominously, Brüning, prodded by Hindenburg, the Reichswehr, and a vast “National Bloc” of industrialists, Junkers, and right wing political groups pushed ahead with the building of the notorious “Panzerkreuzer B,” the second in a series of pocket battleships that were designed to evade the naval limitations of the Versailles Treaty. This step made little sense in terms of either the existing military balance or the Reich’s budget situation. It also made it much harder for even sympathetic SPD leaders to assist the Chancellor, tacitly or otherwise.53 Over time, such moves eroded the basis for cooperative, internationalist strategies. Because the point can easily be misunderstood, it is worth observing that Brüning and his cabinet were not exponents of autarchy, though they were markedly more inclined to compromise on trade issues than earlier cabinets. They did not believe Germany could survive without access to international capital markets and they meant what they said about cutting the budget, consolidating public finances, and retaining access to international credit markets.54 By 1931, however, international trade was shrinking and foreign exchange was becoming scarce even for key currency countries. In both Britain and France interest group pressures to give pride of place in economic policy making to overseas territories and the empire (“imperial preference”) were growing visibly. In France, for example, various business and military groups were noisily organizing a huge exposition for the Spring of 1931 in celebration of the empire’s glories. The efflorescence of such schemes was, in no small part, a reaction to German success in stimulating exports in the midst of world depression. If brought to fruition, they threatened to effectively terminate hopes of running export surpluses. In this context, a policy of Drang nach Osten increasingly looked like either the wave of the future, or, at least, cheap insurance, to growing numbers of German firms, including several giants, such 23
4259 Ch01 4/12/02 9:42 am Page 24
24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
as I. G. Farben (which had close ties to the cabinet and political circles around the Chancellor), Siemens, and Krupp.55 While some early proposals imagined expansion into “Mitteleuropa” under the auspices of a joint Franco-German condominium, with the French contribution consisting mostly of capital, the steady shrinkage of markets turned such proposals into zero-sum games and greatly sharpened national rivalries. By 1931, an eastern-oriented strategy was increasingly emerging as an alternative, not a complement, to traditional internationalist economic policies, not least because some sort of exclusive tariff concession would be needed to ensure Germany’s primacy in these areas.56 The plans for what became the notorious proposal for a customs union between Germany and Austria did not originate with the Reichstag or other normal policy making channels. Instead, the scheme emerged from a secret planning committee convened by the aristocratic German Foreign Office, which had for years met regularly with leading German firms and enjoyed close ties to the military. While Brüning and the Foreign Minister (the project’s public champion, whose reappointment had been confirmed by President Hindenburg) knew of the proposal, the cabinet was not told until the last minute. Advised by Brüning that the timing was unfortunate but that Germany could not control this, the cabinet voted approval on March 18.57 Ever since the end of World War I, the specter of Anschluss between Germany and Austria had alternately tantalized or repelled many Europeans. While the Treaty of Versailles barred Germany from incorporating Austria, the German government’s interest in eventual amalgamation was patent, as was the impatience of much of the Right to see it brought about. Another postwar peace accord, the Treaty of St. Germain, blocked Austria from making any agreement that might threaten its sovereign existence. When news of the plan leaked, a storm of protest immediately blew up. Unimpressed by the legal justifications offered by Germany and Austria, France moved convulsively to counter the German thrust. The chilling implications immediately dawned on the press and financial markets. As the New York Times headlined in late March: “Paris loan market shut by Reich move. Union held to kill prospects for intermediate or long-term credit aid by French. Outcry in press is fatal. Bankers fear customs plan will shatter regained faith of our investors in Europe’s stability.”58 For a few weeks, hope centered on the search for a diplomatic masterstroke that might save the situation. “Financial centers are hoping, therefore, that Aristide Briand will again have recourse to one of those friendly compromises for which he is famous.”59 But Briand, who only a couple of weeks previously in the French Chamber of Deputies had declared the notion of Anschluss dead,
4259 Ch01 4/12/02 9:42 am Page 25
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
25
was seriously embarrassed. As he and the Socialists backpedaled furiously, and the right agitated, French policy shifted decisively. At the same time, the German Right was raising a storm over the opening to the French. Sudden appearances by French industrial groups offering to help bail out hard pressed German cities by buying shares in municipally owned utilities caused consternation. The government’s decision to limit such sales placed it in the position of formally discouraging precisely the capital flow that it needed to survive. On the very eve of the final decision to go ahead with the customs union, an aristocratic acquaintance of Hindenburg’s also complained directly to him about remarks by the French finance minister that seemed to imply that French loans to Germany were intended to perpetuate Germany’s position of military and strategic inferiority.60 The German government sought alternate sources of capital. With British support, Brüning’s government and the Reichsbank renewed pressure on the BIS to live up to its brave talk about leading the way to recovery in central European money markets. But now, ironically, the Germans’ earlier success in reviving international financial confidence helped undo them by driving otherwise sympathetic American bankers into coalition with the French. In late 1930, both the British government and the Governor of the Bank of England, Montagu Norman, had strongly encouraged Brüning as he firmed up his stance on the budget and reparations. After Norman submitted a memorandum to the BIS governing board, a formal proposal for a more active policy soon surfaced, championed by Sir Robert Kindersley, a prominent London private banker serving on the boards both of the Bank of England and the BIS.61 But the revival of Young Plan loans in the late winter emboldened skeptical opponents of BIS action. Their counterattack put Brüning and Luther on the spot. Political support for their stance on the budget was soft. Reichsbank President Luther was in a personal contest with Dr. Schacht for next Presidency [of the Reichsbank] and is staking his chances on a policy of fulfillment of Young Plan which Dr. Schacht opposes largely on ground that the Bank for International Settlements is not stimulating trade and capital movements[.] Dr. Luther and Brüning eager to steal his thunder by pointing to some example of our effort [to] promote capital movements as a justification of government’s change of view and abandonment of moratorium talk . . . [F]lat pronouncement at present that we shall always limit our activities to being a reserve bank specializing in 90 day bills would unloose a volcano.62
Warnings that Luther was “very despondent” and would find it difficult “to keep German papers and banking and political circles from launching radical attacks on [the] Bank,” failed to move a majority of the governing board of the BIS. Citing the dramatic rise in the price of Young Plan bonds, American 25
4259 Ch01 4/12/02 9:42 am Page 26
26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
bankers stood on the principle of laissez faire: “we believe that Germany and other foreign borrowers will, on the whole, get as good credit as they deserve in the market from time to time without the intervention of artificial agencies.” Latin bankers suspected an “English effort to direct investment of French and American funds” while the French resisted repeated Belgian efforts to change their minds. As a result, in the late Spring, as the German financial situation grew increasingly dire, the BIS put off action on a series of proposals for more aggressive intervention in European financial markets advanced by the Belgians with support from Norman and Reichsbank President Luther.63 Closure of the international market for medium and long term loans put Brüning and his government in a box.64 At the very moment that his budgetary successes were turning heads around the financial world, the Chancellor and a handful of his closest collaborators received the dismaying news that their task was Sisyphean and that access to even short term credit would soon be at risk. At the beginning of March, Finance Minister Dietrich quietly warned Brüning that the deficit in the new budget would be much larger than anyone anticipated – perhaps as much as 430 million Reichsmarks – and that the special fund for unemployment compensation required an immediate loan or infusion of another 83 million RM. Dietrich also told the Chancellor that the seasonal flow of funds was creating a short run cash shortage for the government in the next quarter (April to June) of over 400 million RM.65 The news staggered the Chancellor and the head of the Reichsbank, who met with a small group of key ministers to discuss the situation. Issuing a “sharp warning” against any discussion outside the group, Brüning assessed the situation in dire terms. “The German people will not accept any further financial compression without decisive steps on reparations.” The Chancellor insisted, however, that no step in this direction could be taken until he had made “decisive progress” toward restructuring municipal finances and met with German ambassadors at a meeting scheduled for the end of April.66 Brüning persuaded the Reichstag to insert a special provision in the new budget allowing the government to make additional expenditure cuts if revenues continued to fall. This won him further plaudits from the financial community, which did not immediately realize why the provision had been inserted. But when the government quietly sounded out Lee, Higginson about the possibility of floating a new, large loan, the American house gasped when it heard the numbers. While the investment house and the government deliberated, the Germans took advantage of the afterglow of their budget triumph to roll over some 53.5 million RM worth of Treasury bills and sell 71.5 million RM of new T-bills.67 As Lee, Higginson prepared to approach the Banque de France with a proposal to recycle French money into German government T-bills, the
4259 Ch01 4/12/02 9:42 am Page 27
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
27
furor over the customs union exploded. Other German efforts to scout loans proved equally unavailing.68 In the meantime, the deficit projections kept growing. Desperate to avoid repeat performances, the government widened its search for possible shortfalls. By May, the Finance Minister projected a total deficit for all levels of government at 2.2 billion RM.69 But the government found it difficult to settle on a course of action. All alternatives looked disastrous. The Chancellor was convinced that a bid for complete cancellation of reparations would have to await, not only the upcoming French Presidential election and the long scheduled Disarmament Conference, but also the outcome of the 1932 American elections. He was also sure that, as bad as things were now, 1932 promised to be worse, and that the BIS would not help.70 But he also believed that if he announced yet another austerity program (only realistically possible through Notverordung) to tide the government over the rest of the current year, “the existence of the present Cabinet is in danger, if one does not seize upon (anschneide) the reparations question.” At the same time, however, Brüning believed that the fallout from a unilateral halt to reparations would be so terrible that the step was essentially unthinkable. Indeed, the protocols of their meeting suggests that the Chancellor and his ministers were not willing to surrender power to a government that would try it.71 The Foreign Minister and the Reichsbank President reinforced these forebodings. Both warned vehemently against any steps that might reawaken foreign anxieties about Germany’s commitment to paying reparations. Brüning resolved his dilemma by a kind of “doublethink.” The Chancellor finally concluded that only a truly desperate expedient offered any hope of relief:
26 27 28 29
[T]wo things are necessary: From an internal standpoint, release of the [new] Notverordung must awake in the people the impression that the revision [of reparations] has already begun; abroad, by contrast, the impression must be awakened that we are making all exertions to fulfill the plan.72
30 31 32 33 34 35 36 37 38 39 40
The Chancellor’s plan for bringing off this quixotic program rested on the combination of a diplomatic offensive with the new, draconian Notverordung. Aware that Britain had the biggest direct commercial interests in Germany and central Europe, he hoped to jolt the British into realizing that Germany could no longer afford to pay reparations by issuing a shockingly austere Notverordung on the eve of his scheduled talks with British Prime Minister McDonald at Chequers.73 But the late Spring of 1931 was not a propitious time to try to fool all of the people all of the time both in and out of Germany. Though some historians have maintained that the depths of the German budgetary crisis only emerged 27
4259 Ch01 4/12/02 9:42 am Page 28
28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
on June 9, when it became clear that tax revenues for April and May had run far below projections, in fact the government’s predicament had become apparent long before then.74 Some cities were near default or actually in arrears, forcing the Reich to step in in several instances. Since municipal loans in particular comprised a major share of the portfolios of some smaller municipal savings banks, this stirred some fears of local defaults and put additional pressures on Reich finances.75 Official publications of the republic’s monthly tax revenues and a variety of budget numbers were regularly scheduled and closely watched. Readers of the Frankfurter Zeitung, probably the leading newspaper in Germany, for example, could read virtually the same figures that economic historians now use to illustrate the size of the tax shortfall soon after they were compiled and published by the government.76 In the unlikely event that readers failed to grasp the meaning of the numbers, the paper’s increasingly alarmed articles on the growing budget shortfall could hardly have failed to alert them, or the financial community that watched anxiously as the Young Plan bonds tanked. Not surprisingly, after hitting their respective year’s highs in April (see Table 4), both the German stock and bond markets began steadily selling off, long before news of the Creditanstalt debacle arrived on May 11.77 In early May, well before news of the Creditanstalt burst upon the world, the German government attempted to float a domestic bond issue for the Reichspost. Only part of the loan could be sold, generating embarrassing publicity. This debacle plainly owed nothing to international currency pressures and everything to fears by domestic investors of heavy losses.78 By mid-May discussions between the political parties and the government over how to close the budget gap were becoming very tense. Paramilitary clashes, already running at high levels, intensified, while the national SPD leadership, not coincidentally, voiced doubts about spending cuts of the size Brüning thought was necessary. Leaders of the Prussian SPD, however, who carried enormous weight both within the party and the government, faced elections in early 1932. They spoke out in favor of making big cuts now, rather than risk having to go through the whole process again later in the year. As Brüning and the cabinet – who still lived in the shadows of 1918, the Kapp Putsch, and the myriad other disorders of early twenties German politics – geared up for the crunch, the crack of doom resounded: the Minister of the Interior warned that he could no longer guarantee the maintenance of order.79 Well aware, as some historians seem to have forgotten, that he would need votes from the right as well as the left to sustain his budget when it faced the inevitable challenge in the Reichstag, Brüning appears to have gradually shortened his time horizon, from getting all the way through to the beginning
4259 Ch01 4/12/02 9:42 am Page 29
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
29
of 1932 to just getting past the Chequers meetings.80 His language about reparations also became increasingly direct and reckless, though to the very end he did not want to admit that he was more than on the verge of repudiating them. By contrast, his Finance Minister went considerably further, in private threatening resignation if the Chancellor foreswore the right to declare a moratorium in the meetings with the British and call public attention to the need to revise reparations.81 By then the climate of both German domestic politics and European diplomacy were garishly different from the sunny optimism of early March. “Systemic risk” was rising rapidly. Just days before the German budget crisis came to a head, Aristide Briand narrowly lost his bid for the French Presidency. While factors internal to French politics surely figured in this outcome, the customs union proposal had helped torpedo Briand’s campaign. The Stahlhelm staged another in their series of gigantic rallies at which the French and their allies, the Poles, as well as, of course, “Marxists” within Germany, were roundly denounced.82 On May 19th, the launch of the pocket battleship Deutschland kicked off a new round of competitive naval construction between France, Germany, and Italy.83 On June 3, at a meeting of the celebrated Langnamverein, a trade association of heavy industrialists, including many from the Ruhr, Albert Vögler, a leading figure in the steel industry, demanded the cession of all payments of foreign “tribute” and a shift in policy toward building up the home market. General Gröner, who along with his then protégé, General Schleicher, provided critical support for the government within the Reichswehr and Hindenburg’s circle, also advised Brüning to wait no longer on reparations. This was advice that the Chancellor dared not ignore.84 For essentially domestic political reasons, then, Brüning made the fatal decision to denounce reparations and proclaim that Germany had paid all that it could. The Notverordung that Hindenburg signed on June 5 and was published, together with an accompanying manifesto, the next day stunned not only Germany but the world.85 Since at least May 25, German newspapers had been speculating that Brüning was likely to ask for some sort of relief in regard to reparations. The French government and, doubtless, others, had drawn the inevitable conclusions. Suspicions that Germany might postpone reparations transfers had been circulating for some time, fanning the run on German gold reserves and major banks that appears to have begun about May 25. On the 28th of May, Young Plan loans fell sharply all over the world, “on a rumor that the German government were contemplating the declaration of a reparations moratorium.”86 The distinction that Brüning and other German officials planned to highlight outside of Germany – that the official announcement did not formally repudiate 29
4259 Ch01 4/12/02 9:42 am Page 30
30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
reparations or invoke the Young Plan’s provisions for postponing payments – no longer mattered by the time of publication. The manifesto’s sensational references to “tributary payments”and “intolerable reparation obligations” as well as its provocative language about a people’s capacity for suffering having reached its limits, spoke in stentorian tones to investors in and out of Germany.87 Read everywhere, it was everywhere read as heralding a broader German inability to make international payments, not least because it was accompanied by official denials that Germany would soon be forced to suspend payments on both reparations and private debts.88 Investors converted Reichsmark deposits into gold, as the Reichsbank’s gold cover suddenly began to look dangerously thin. Reichsbank gold registered its first decline of the year in the week ending June 6th. This can be seen clearly from Table 8. Gold reserves did not fall in March when the customs union was announced. Neither did they fall after the Creditanstalt collapsed – indeed, both Reichsbank gold and total foreign exchange reserves rose all through the month of May, though the entries for May 30th probably reflect a final payment the government received as part of a deal struck in 1929 for a match monopoly.89 In the two weeks after June 6th, Reichsbank gold reserves declined precipitously – by 40% – before stabilizing briefly at 60% of their previous level as it became clear that the Reichstag would not overturn the Notverordung. Fears for the stability of the Reichsmark were also reflected in the premium for gold bonds over currency bonds, as shown in Table 4. Comparing the relative prices of the 6% bonds reveals that the gold premium calculated from the ratio of bond prices drifted down from 6.6% at the start of the year to 6.1% in April. It rose to 6.9% in May, and then soared to 10.0 in June, as investors became very anxious indeed about the stability of the Reichsmark. Underlying the run was the wrenching fear that the government could not make its colossal budget cuts stick. In the early days of the panic, with Brüning in England for the bilateral talks, this seemed likely. Across the political spectrum, the new budget package aroused “general horror.”90 The SPD, the parties of the Right, and the Stahlhelm (which had been afforded a share of the administration of youth programs in the budget) all vociferously denounced the cuts. Göring and other Nazis furiously attacked them. On June 11, Stresemann’s old party, the DVP, split thunderously over the cuts. Since the party was somewhat inaccurately still labeled as uniquely a party of big business, the split was widely noted, even in foreign business centers. By June 13th, the gold drain was so severe that the Reichsbank raised its discount rates two full points in an attempt to sustain the gold value of the Reichsmark. Over the next few days, the financial situation worsened,
4259 Ch01 4/12/02 9:42 am Page 31
German Currency Crisis
Table 8.
1
31
Weekly Reserves at the Reichsbank (Million RM).
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
Date
Gold
Foreign Exchange
Silver
7.1.31 13.1.31 23.1.31 31.1.31 7.2.31 14.2.31 23.2.31 28.2.31 7.3.31 14.3.31 23.3.31 31.3.31 7.4.31 15.4.31 23.4.31 30.4.31 7.5.31 15.5.31 23.5.31 30.5.31 6.6.31 15.6.31 23.6.31 30.6.31 7.7.31 15.7.31 22.7.31 31.7.31
2216 2216 2244 2244 2244 2254 2266 2285 2285 2286 2286 2323 2344 2345 2348 2368 2370 2370 2370 2390 2300 1766 1411 1421 1422 1366 1353 1368
400 268 196 199 198 181 175 166 189 209 223 188 166 114 132 158 169 171 197 186 113 104 93 300 371 244 160 246
161 200 207 172 178 192 202 160 166 179 196 157 143 187 207 168 175 186 200 174 177 199 214 78 84 79 74 45
Source: Die Bank, 1931, various issues.
29 30 31 32 33 34 35 36 37 38 39 40
exacerbated by the increase in the cost of credit imposed by the Reichsbank. SPD leaders and General Gröner, the Defense Minister, claimed to see revolution around the corner. Stories – which appear to have a foundation in fact – that the military was preparing to intervene swept Berlin and other major cities. Brüning’s own day calendar (Tageszettel) for June 13 indicates that he met with General Schleicher to discuss Reichwehr plans for coping with civil unrest. Insiders were saying that the government was two weeks away from defaulting even on pensions.91 31
4259 Ch01 4/12/02 9:42 am Page 32
32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
All the tumult, along with frantic bargaining by Brüning (whose train, as he returned from Britain across Germany, was mobbed by Nazi thugs and other paramilitary groups) began to change the minds of the deputies. Pressure from various business groups and the Reichsbank intensified. Brüning and the cabinet made it clear that if the Notverordung were overturned, they would all resign. Once again, Brüning threatened to overturn the SPD-led coalition that ruled Prussia. The chancellor also offered a few, minor compromises.92 Economic historians who focus on the Creditanstalt and international factors in the banking crisis have noticed that something happened on June 16 that brought an abrupt halt to the foreign runs on German banks and the Mark. But they have typically identified it as the decision by Montagu Norman, Governor of the Bank of England, to float an emergency loan to the Austrian National Bank. No doubt this decision helped stabilize Austria, but why should a British decision to bail out Austria have stopped internal drains from local German banks? Our answer is, of course, that it did not. Rather, something else that happened the same day that stabilized both domestic and international markets in which the Mark figured: The SPD and other parties decided not to try to overturn the budget. Brüning had won, and the deputies dispersed (Winkler, 1993, p. 413). This political success brought Brüning only a few days respite. On the heels of the dispersal of the Reichstag came the first, muffled reports of the Nordwolle’s losses. The run on the Reichsbank swiftly resumed, as the shadow of death fell over the giant Danatbank, which was closely associated with the huge textile concern. The Reichbank’s action to defend the mark had increased the cost of borrowing, intensifying the tottering firm’s problems. In ordinary times, the bank could have turned to the Reichsbank for help. But the battles over the custom union, the budget, and Brüning’s ill-fated references to reparations “tribute” had left the Reichsbank in no condition to act as lender of last resort. As late as the end of May, the Reichsbank’s cover of gold and foreign exchange exceeded 60% of the value of its own notes and 47% of the money supply as a whole. The gold cover fell dramatically in June, as shown in Table 8, threatening to breach the legal limit as word of the Nordwolle’s condition began to leak out. The Reichsbank could no longer act as a central bank to its domestic constituents.93 This, of course, is what happens in countries on a gold standard. The central bank focuses on maintaining the value of the currency. It needs to buy and sell gold in response to international demands; it cannot divert its attention to care for commercial banks. The Reichsbank was a textbook example of this condition. It had maintained the value of the Reichsmark as its principal domestic constituencies and the Allies concerned about reparations payments
4259 Ch01 4/12/02 9:42 am Page 33
German Currency Crisis 1 2 3 4 5 6 7 8 9
33
desired. But it could not at the same time take care of problems within Germany. When the force of the depression caused a few large firms to default on their loans, banks clearly needed some help. The Reichsbank was unable to help and the banks failed. But the problems did not originate in bank behavior; they arose from the policy stance of the Reichsbank and from the political acts of the government that put its attention to that policy to the test.94 Grossman (1994) showed that no country that went off gold experienced a banking crisis. If Brüning had controlled trading in the Reichsmark earlier or announced the customs union later, German banks probably would not have failed.
10 11 12
VI. CONCLUSION
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
Keynes said during a discussion of the Young Plan in July 1929 that, “even if it were foreseen, the announcing of a moratorium would so damage Germany’s credit that a crisis would be precipitated by the announcement.”95 He was right. Brüning’s announcement that Germany could pay no more reparations generated a run on the mark in which banks failed and currency transactions were controlled. The ensuing run was on the Mark, not banks, because the fear was that the German governments, not the Grossbanken, would not be able to pay their bills. Could the banking crisis have been avoided? Almost certainly. Brüning abrogated his international obligations in three ways during the spring and summer of 1931. He promoted the customs union, geared up to stop reparations payments, and imposed currency controls. He did them in this order, placing great strain on banks. Had he done these three acts in any almost other order, German banks might well not have failed. Had Brüning gone off gold by imposing currency controls before one of the other announcements, German banks probably would have survived.96 Could the currency crisis have been avoided? That is a harder question. Adherence to the gold standard caused problems for countries far more stable than the Weimar Republic. It is unlikely that Brüning could have avoided going off gold in the end. Had he thought ahead, it could have been done more gracefully and with less damage to the German economy. We therefore conclude that the German crisis of 1931 was a twin crisis caused by a run on the currency triggered by domestic political actions. Banks failed, and the international economy constrained Germany, but the primary actions that led to the crisis of July 1931 were German policy initiatives, taken by Brüning in response to political pressures facing him and his supporters. The German crisis of 1931 was made in Germany. 33
4259 Ch01 4/12/02 9:42 am Page 34
34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
THOMAS FERGUSON AND PETER TEMIN
We draw another lesson as well from this analysis: The terms of the celebrated debate over the so-called “Borchardt thesis” need to be recast. Recent work showing that political forces in favor of reflation did exist inside the Weimar regime has made an important contribution.97 But it is time to acknowledge that the truly significant political coalition that failed by only a hair’s breadth was transnational. Put simply, there were powerful political forces in both France and Germany that perceived the advantages of an internationally cooperative strategy. This cross-national coalition always faced major obstacles, but as late as the Spring of 1931, its failure was not obvious. Indeed, the cause of Franco-German cooperation, if not reconciliation, appeared to be making great strides forward. What destroyed it was the German government’s decision to pursue Drang nach Osten. It is all very well to say, as Duisberg and other apologists for the policy did, that this decision was a response to pressures for “imperial preference” among the allies. So it was. But the reckless pursuit of “Mitteleuropa” destroyed the possibility of French financing for Germany. Stabilizing Germany by this means was clearly a goal of Briand’s policy, and he came heartbreakingly close to bringing it off. What Borchardt and his followers now owe the rest of us is an answer to one simple question: Are they prepared to argue that German business groups and policy makers also had “no choice,” no “Handlungsspielraum” in this decision to swallow central Europe? That is what the denial of the possibility of an international version of reflation, if not full throated “Keynesianism” requires. As late as March, 1931, this seemed not a remote contingency, but a reasonable bet – one that even bankers were willing to take.
24 25
NOTES
26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
1. For a good inventory of the main works in both English and German, see Balderston (1994, p. 43, n. 1). 2. See, e.g. the discussion of the German reaction in James (1986, p. 291) or Born (1967); also Balderston (1994). 3. James (2001a, p.129). See also p. 16. By contrast, the emphases in James (2001b, pp. 59–63) more closely parallel his 1986 study. 4. See the discussion in Eichengreen and Hatton (1988) and Stachura (1986). Unemployment figures for Germany were published monthly in the German Statistisches Reichsamt’s widely read Wirtschaft und Statistik. The same figures also appeared in the Monthly Bulletin of Statistics published by the League of Nations, where it is explained that they are for the end of the month and cover trade unionists. These are the figures we use here. Regardless of how one might want to adjust them in the light of better information about coverage or the labor force, they are useful as an indicator of monthly trends that could actually be tracked by contemporaries. 5. The literature on the Depression’s effects on Germany is immense; besides the sources we cite for our statistics and those cited in note 1, above; see, e.g. Winkler (1993) or Feuchtwanger (1995).
4259 Ch01 4/12/02 9:42 am Page 35
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
35
6. Feuchtwanger (1995, pp. 70–71, 219–220); Winkler (1993, pp. 369ff.); Kolb (1993, p. 214). For the role played by I. G. Farben and German heavy industry in the change, see the account of the circumstances surrounding Duisberg’s letter to Finance Minister Moldenhauer in Winkler (1993, pp. 369–371). Duisberg headed the famous Reichsverband der Deutschen Industrie, the leading (but not the only) big business association in Germany. He was also a top official of the giant chemical company. The RDI had begun pressing for the fall of the government as soon as the Young Plan cleared the parliament. Plumpe (1990, pp. 529–532) offered a critique of an earlier Winkler essay on the changeover. This controversy has no bearing on our paper, but see the discussion of the dissolution of the cabinet (which does not mention Plumpe) in Winkler (1993, p. 371). 7. Kolb (1988, pp. 114ff.; the quotation appears on p. 114); his later German rendition is entirely consonant with this judgment. The discussion in Winkler (1993) is also revealing. 8. See, along with Table 2, Reichsamt (1933c, pp. 341–342). These figures underestimate the total debt, excluding, for example, another 4 billion RM borrowed by public corporations in their own names. See Reichsamt (1933c, pp. 345–346). (Here “billion” is used in the American sense of 1000 million.) 9. Estimates of total German indebtedness vary, though the differences do not add up to much for our discussion. Early German estimates were sometimes suspect, since they were prepared as part of negotiations over war debts and reparations. But the official German estimates made in the late twenties and after, appear quite good and have hardly changed at all over the years. Compare the figures in Reichsamt (1933c, pp. 341–350) with those in Länderrat des Amerikanischen Besatzungsgebiets (1949, Vol. II, p. 354) and Deutsche Bundesbank (1976, p. 313). See also the helpful discussions in Bachmann (1996, pp. 42ff.) and Thomas (1934). The 1931 Wiggin Committee’s report, sometimes referred to as the “Layton Report” after Sir Walter Layton, who played a major role in its drafting, appears to have been the source for much of the data published during this period in The Economist (which Layton edited). These two sources formed the basis for the estimates in Feinstein and Watson (1995, p. 127). Note that debts of public corporations were reckoned separately from the rest of the official debt totals, though they also borrowed through other governmental agencies; see Reichsamt (1933c, pp. 345–346). It is also important to recall that all through the Weimar period, a low intensity war raged in the courts and parliament over the possible redemption of older public debts that had been inflated away; cf., Southern (1981). Finally as emphasized by McNeil (1986), truly remarkable efforts were often undertaken to hide borrowing or evade the (usually weak) German or American efforts to restrict borrowing. 10. The literature is immense; see, Winkler (1993, pp. 380ff.); Feuchtwanger (1995, pp. 222–224). The political developments came in the context of sometimes rapidly changing strategies by major industrial and agricultural groups. See, for example, the literature discussed in Kolb (1993) or Abraham (1986). The Finance Minister, a former I. G. Farben executive, resigned on June 18. The budget package failed in mid-July. Not all spending was cut; funds for so-called Osthilfe (agricultural aid tilted heavily toward Junkers) increased. Communities were also authorized to tax beer. See Feuchtwanger, 1995, p. 224. 11. The one-third estimate does not include the 41 deputies of Hugenberg’s Deutschnationale Volkspartei, who really constituted an anti-system party of the far right. 35
4259 Ch01 4/12/02 9:42 am Page 36
36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
The Reichstag then had 577 members. The practical effect was to increase still further Brüning’s dependence on Hindenburg. Cf. Feuchtwanger (1995, p. 233). 12. See, e.g. Ellis (1941, p.162). All analyses of the period feature a discussion of the capital outflow. 13. For the “seamless” projection of the1930 crisis, see, for example, Harris (1935) or less baldly, Ellis (1941); a variant is partly explicit and partly implicit in James (1986), though he is not entirely consistent, as noted below in the discussion of the French role. Ritschl (1998) argues that the Young Plan terminated capital flows into Germany by making private debts junior to reparations. In our view this confuses the cart with the horse. Priority became pressing once capital inflows ceased. The best disproof of the hypothesis that Germany could not borrow in the face of the Young Plan loan is the simple fact that it did so, repeatedly, sometimes in the face of warnings from authorities who went out of their way to remind potential creditors of precisely Ritschl’s point. See, for example, the Agent General for Reparation’s comments to the U.S. Secretary of State about the German loan floated by Lee, Higginson on the eve of the public offering of the Young loan in 1930 in U.S. Department of State (1945, Vol. III, pp. 96–102). Cf., also the correspondence engendered by the release of this volume in 1945. This can be found in Box 121 of the Thomas W. Lamont Collection, Baker Library, Harvard Business School. In a letter to Lamont of July 27, 1945, Russell Leffingwell of Morgan ascribed the State Department’s unwillingness to discourage the loan to intervention by Jerome Greene of Lee, Higginson. Lamont, in a draft letter of Aug. 7, 1945 to Arthur Hays Sulzberger of the New York Times, noted that Greene and the Under Secretary of State, Joseph Cotton, who played a key role in the episode, had been classmates at Harvard. That draft was not actually sent; other correspondence appears to have been substituted for it. For this paper, the important point is that major investors were well aware of the Young loan and the priority issue; as we show, they kept loaning anyway. 14. For Lee, Higginson, see, e.g. Ellis (1941, p.162) and Bachmann (1996, pp. 213ff.), though we cannot follow her judgments on several broader issues. As discussed below, the Thomas H. McKittrick Collection, now at Baker Library, Harvard Business School, contains important letters and notes concerning the Lee, Higginson deals with Germany. For the post-election maneuvers on the budget, see, again, Winkler (1993, pp. 394–398) and Feuchtwanger (1995, pp. 236–237). 15. On Prussia, the Notverordung, and the Social Democrats, see the discussion in Winkler (1993, pp. 392–402) and Feuchtwanger (1995, pp. 233–238). 16. The standard, and generally excellent, source for discussions of reparations revision is Bennett (1962). But our account, presented in more detail below, of the BIS, Norman, and German government’s stances on the “Kindersley Plan” draws on documents that were not available to him. For the British government’s efforts to influence Brüning, see Sir H. Rumbold (British Ambassador to Germany) to Mr. A. Henderson (Foreign Minister), March 4, 1931, in Woodward and Butler (1947, Vol. I, p. 573). While the letter is dated March, it discusses events that occurred in December, 1930 and January, 1931. For the campaign against the BIS, see for Schacht, e.g, New York Times, Jan. 2, 1931, p. 15; for Georg Solmssen, a Deutsche Bank director serving as President of the German Banker’s Federation, see Commercial and Financial Chronicle, Jan. 3, 1931, p. 51. In December, 1930, the Swiss government ordered the Basle head of the (German)
4259 Ch01 4/12/02 9:42 am Page 37
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
37
Allgemeine Pressebureau to leave the country after repeated French complaints that he was agitating against the BIS. See the letter of the French consul at Basle to M. de Marcilly, the French Ambassador to Switzerland, Dec. 12, 1930 in Carton B32221 in the Archives of the French Ministry of Finance at Savigny-le-Temple, France. The German journalist was said by the consul to be continuing the campaign from just across the border. 17. The quotation is again from Ambassador Rumbold’s letter to Foreign Minister Henderson, March 4, 1931, in Woodward and Butler (1947, Vol. I, pp. 576–577); and refers to the time period we discuss here. See also New York Federal Reserve Governor George Harrison’s letter of Feb. 9, 1931 to Norman Armour, which explains that “the best opinion” was that “Germany is unlikely to declare a postponement of transfers,” New York Federal Reserve Bank Archives, C261, Germany-German Gov’t 1916–1950. A contemporary account sensitive to the nuances of the reparations issue in these weeks is Dulles (1932, p. 234). The Germans believed that over time their foreign creditors would come to perceive that their own interests lay in relenting, either because they feared social chaos and the drift to the right in Germany itself or because their own industries found it impossible to compete with a major competitor whose price level was falling. See, e.g. Kolb (1993, p. 218). Brüning saw his policy of cutting the pay of German government officials as part of this strategy of general deflation. Cf. James (1986, p. 68). The wage cutting strategy was strongly supported by many industrial groups. 18. Sir H. Rumbold (British Ambassador to Germany) to Mr. A. Henderson (Foreign Minister), March 4, 1931, in Woodward and Butler (1947, Vol. I, pp. 573–574). Rumbold was generally a pessimistic observer of the German scene. 19. For the Institute of International Finance, see New York Times, Jan. 20, 1931, p. 35 and Institute of International Finance (1931). The Commercial and Financial Chronicle, Jan. 31, pp. 763–764, outlines the study’s conclusions and notes that the Institute was partially supported by the Investment Bankers Association of America. For the British government, see Mr. A. Henderson to Sir H. Rumbold, Feb. 19, 1931, in Woodward and Butler (1947, Vol. I, p. 559). For the BIS see, besides our discussion below, “Remarks of the President of the Bank for International Settlement Before the American Club in Paris,” Feb. 12, 1931. This was sent to Thomas Lamont by Leon Fraser, then of the BIS. See also N. D. Jay to Lamont, Feb. 14, 1931, which encloses a memo by Montagu Norman that had been discussed at a BIS meeting. All these are in Box 181 of the Thomas W. Lamont Collection, Baker Library, Harvard Business School. See also Dulles (1932, p. 218) and the comments below on the BIS. Her study (p. 224) is again sensitive to the improvement in the German position in this period – perhaps not surprisingly, since she was the sister of John Foster Dulles, who had been the American expert on reparations at Versailles and was then generally recognized as the leading American authority on the issue. See Ferguson (1995, p. 144). 20. The first two quotations come from The Economist of Jan. 3, 1931, pp. 17–18. The discussion there merits a longer look: “After the marked decline in deposits shown in the October bank balance sheets, in consequence of the credit crisis, the November balance sheets show every sign of returning calm . . .. The cessation of the withdrawals of money, especially foreign withdrawals, has enabled the banks to reduce their rediscounts at the Reichsbank, which had risen during October . . ..” The final quotation is from a cable from Morgan et Cie., to J. P. Morgan & Co. in New York of Jan. 12, 1931, The Pierpont Morgan Library, New York, Morgan Bank European Papers, Box 13. 37
4259 Ch01 4/12/02 9:42 am Page 38
38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
21. For the Morgan bank exchange, see the cable from Morgan et Cie., to J. P. Morgan & Co. in New York of Jan. 12, 1931, The Pierpont Morgan Library, New York, Morgan Bank European Papers, Box 13. For the Reichsbank exchange with the New York Fed, see the former’s cable to the Fed of March 11, 1931: “Offers of foreign money are said to be more numerous,” while “flowed out German capital stimulated by the rise of shares, seems returning to Germany,” New York Federal Reserve Archives, C261, Germany-Government 1916–1950. The translation is, again, clumsy but clear. 22. The reparations payments had been deliberately scaled down and made to vary over time under the Young Plan. See, e.g. Moulton and Pasvolsky, War Debts, pp. 223ff. 23. New York Times, Feb. 9, 1931; the comment occurs in a story about the “sharp advance this week” of Young Plan bonds and is clearly meant as a broader generalization. Cf. also Dulles (1932, p. 225): “The Bank [of International Settlements] was in a position to explain the creditor point of view to Berlin – the view that Germany’s balance of trade was likely to improve, and that her main difficulties were caused by internal politics.” 24. For long rates, see the tables on pp. 128–129 of Institut für Konjunkturforschung (1933). Note that quotations for the Young Bonds for the end of June were affected by the brief impact of the Hoover Moratorium. See the discussion below. Then and later, observers argued over the extent to which some of these improvements were seasonal. Completely consonant with our analysis of the key role of the budget, between January and May, industrial long rates fell considerably more than “public loans” or “communal [city] obligations.” But all fell, though we know of no conclusive data on the quantity of lending at these rates. It is a commonplace that many large German banks bought their own stocks in the wake of the 1930 crisis. See, e.g. Born (1967, p. 60). But they cannot have been buying the broader market, which was also rising. Bank stocks rose along the rest of the market in January, February, and March, with the difference in April being too small to make anything of. Born’s observation that the Berliner Handelsgesellschaft succeeded in privately placing shares in New York in January is consistent with our view of the external perceptions of the German economy then. For modern attempts to predict currency crises, see Goldstein, Kaminsky and Reinhart (2000); their discussion of potential indicators in Chapters 2–4 can be instructively compared with the available German data. 25. Authoritative contemporary accounts, such as Toynbee (1932, p. 518), reported the amount we use here; Bachmann (1996, p. 242), suggests the final amount totaled about RM 134, 400,000. The New York Times, Feb. 6, 1931, p. 11, reported a dollar figure of $32 million. The German state’s demonstrated ability to borrow made an impression on contemporaries, see, e.g. Rosenstiel (1931a, p. 18). As we discuss below, Bachmann’s determinedly pessimistic account notes the borrowing, without, in our view, absorbing its implications. Facts come through that are inconsistent with her story line: Germany borrows, bankers are competing to bring out loans, one bank is striving to become the exclusive banker for the government, etc. See Bachmann (1996, pp. 235ff.). It is only fair to note that one banking house queried by the government replied that additional loans would have to await a new reparations settlement that eliminated the possibility that in a crisis recent creditors would have to stand in line behind holders of reparations bonds This is the one case we have found that fits Ritschl’s argument, discussed above. See James (1986, p. 72); the comment came right as Lee, Higginson was floating the loan and, very likely the bank in question realized it was not going to get the business it, too, had scouted.
4259 Ch01 4/12/02 9:42 am Page 39
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
39
26. New York Times, March 30, 1931, p. 30; this article is effusive about Brüning’s success. For the short term debt reduction, see Bachmann (1996, p. 254), who recognizes that the improvement runs counter to her own pessimistic appraisal of the Reich’s credit situation. 27. J. P. Morgan, Thomas W. Lamont, and S. P. Gilbert of J. P. Morgan & Co., New York to their Paris partners at Morgan et Cie., for relay to Gates McGarrah of the BIS, March 14, 1931. The copy we draw upon is in The Pierpont Morgan Library, New York, Morgan Bank European Papers, Box 13. For more on the context, see below. Even U.S. Ambassador to Germany Frederic Sackett, a famously consistent bear on prospects for Germany – see Burke (1994, passim) – changed his tune. See Sackett to George Harrison, March 17, 1931, New York Federal Reserve Bank, C261 GermanyGerman Government 1916–1950. The ambassador relates his discussions with Luther and other prominent Germans, who are “impressed with the idea that they have got some way to raise German credit rating in the financial markets and eventually relieve themselves from the immediate pressures of short term debts.” The former Great Pessimist continues, “I have the distinct impression that things are looking better in Germany,” and that “international political moves are reassuring.” As a consequence, “a very considerable amount of German money which fled abroad is returning here, seeking the increased rates which are available.” 28. This is an excerpt from a letter in the Thomas W. Lamont Collection, Baker Library, Harvard Business School, dated March 10, 1931. Both its style and contextual evidence indicate that the author is Lamont himself. There is, however, some only partially decipherable handwriting on a page that raises a slight possibility that the author was Lamont’s Secretary, de Sanchez. But he was ill, probably out of the office, and scarcely in a position to make any wagers. The letter is in any case striking evidence of the shift in sentiment. 29. New York Times, Feb. 6, 1931, p. 11, relates details of French loan participation; on the French reluctance to loan, see, e.g. Bachmann (1996, pp. 216ff.); Bennett (1962) is an excellent general account, though he overestimates the willingness of the French to consider a loan in the Fall. Knipping (1987, pp. 198ff.), by contrast, correctly links the later French shift to the change in the French government. Note that the banks almost certainly did not believe they would be immediately able to resell all the bonds to the public, for they were still holding at least some bonds from the Fall issue; the point is that they were willing to buy them all in the context of what they perceived as an improving climate. 30. On the change of government, cf. Knipping (1987, pp. 198–199). Lee, Higginson had been targeting increased business in France since at least the fall. See, e.g. Paul Courtney to Partners, Sept. 8, 1931. This and subsequent letters in the McKittrick Collection discuss the French banker, René Charron, and the firm’s hopes for FrenchGerman detente. The bank also weighed sending some legal work to Sullivan and Cromwell in New York, in order to enlist the aid of John Foster Dulles. See the telegram to Courtney, signed “George” (Murnane), Sept. 18, 1930, Carton 6, Thomas H. McKittrick Collection, Baker Library, Harvard Business School; on Reel 11 of the microfilm of the collection. It is clear that Dulles and Murnane subsequently worked closely together, but we have no other evidence concerning this episode. See Aalders and Wiebes (1996). For Laval and the Belgians, see the discussion in Bussière (1992, pp. 397ff.), which demonstrates that Laval and the Belgian banker Francqui were communicating since at least late November, 1930. The Belgian motivation seems clear: it is a variant 39
4259 Ch01 4/12/02 9:42 am Page 40
40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
on the answer to the old joke about what mice do when two elephants are becoming restless around them. For Lee, Higginson and Charron, see the correspondence in Carton 6 of the McKittrick Collection, Series II. Both the originals and a microfilm of this collection were consulted. Glashagen (1980, pp. 355) mentions that even the Reichswehr was then interested in exploring a possible (limited) arms accord with the French. But our discussion below relies on many sources not available to him. 31. On the French shift, see the memos from the French Ambassador to Berlin to the French Finance Minister, Jan. 15 and Jan. 29, 1931. These are now in the Archives of the Ministry of Finance, Carton B31478. See also the brief note of Feb. 10, 1931 on the discussions between Governor Moret of the Banque of France and Reichsbank President Luther in Carton B31469 in the same archive. (Unless otherwise mentioned, the Ministry’s archives were still located at Bercy, Paris when we consulted them.) The loan was sharply criticized by the right, but the cabinet voted approval. See New York Times, Feb. 8, 1931, p. 8. For the reaction of financial markets to news of the French participation, see New York Times, Feb. 6, p. 11, which includes the cited passage and mentions that other French banks were participating and Feb. 9, p. 30, which notes the immediate change in sentiment in the Berlin money market. While Briand was clearly a moving force within the French government behind the French shift, French banking opinion had also changed. See New York Times, Feb. 8, 1931, p. 8. Note that Laval formally took the reins of power at the end of January, but the new government had been in the works for some weeks. The American embassies in Paris and Berlin appear to have been surprised by the switch and somewhat in the dark about its meaning. See, the discussion in Burke (1994, p. 115) and the full texts of the key cable he mentions, Edge to Secretary of State, Jan. 29, 1931, reproduced on Reel 34 of (U.S. Department of State, 1986). It seems clear that Murnane withheld all mention of a possible French opening when he saw Sackett in January; we also think that Murnane’s comments at that time about how Lee, Higginson rather than Germany was key in selling bonds were very much comments of someone striving to bring off an exclusive arrangement with the German government. See Burke (1994, p. 114) and the full text of Sackett to Secretary of State, Jan. 21, 1931, also on Reel 34 of U.S. Department of State (1986). 32. The quotation is from Dulles (1932, p. 236), who also references McGarrah’s speech. This is the speech in the Lamont Collection, mentioned earlier. The public discussion of gold and recycling appears in, e.g. New York Times, editorial, Feb. 5, 1931, p. 22 and Feb. 9, 1931, p. 30. The discussion was much broader, however, and linked to discussions about gold distribution (e.g. The Economist, Jan. 3, 1931, p. 7). Essentially a cliché within financial circles at the time, recycling was a basic premise of the many efforts by Lee, Higginson and others made to promote rapprochement between France and Germany. Heyde (1998, pp.142–143) notes the interest of some prominent French industrialists in the developing discussions with the Germans, though his broader, more skeptical, narrative does not take sufficient account of the financial evidence we discuss here. 33. James (1986); the most detailed is Bachmann (1996), whose work is so careful and painstaking that it records a great amount of detail that is inconsistent with its central theses, as we note. Both briefly discuss the French episode, but neither pursue its implications systematically. In effect they assume that the breakdown of relations between France and Germany was inevitable, without making much of an argument.
4259 Ch01 4/12/02 9:42 am Page 41
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
41
34. Luther had long been close to the firm and owed his appointment as Schacht’s successor at the Reichsbank in part to a dramatic intervention by the bank’s top executive. See Treue (1972, pp. 64–65), which also has an excellent overall account of the firm’s place in German banking history. Bratz (1958, pp.10–11) indicates that Mendelssohn was one of the handful of private banks that could count on state help in case of trouble. For Lee, Higginson’s relation to Mendelssohn, see the papers in Box 6 of Series III of the Thomas H. McKittrick Collection, Baker Library, Harvard Business School. As one Lee, Higginson telegram to “Courtney” in Paris (unsigned but almost certainly by George Murnane, of Sept. 17, 1930) summed up: “Consider our strongest reliance is Schaeffer who is loyal to us solely through Kempner [of Mendelssohn].” The American bank’s anxieties about competitors are obvious in the correspondence referred to above; at least one rival is mentioned there. But reading Bachmann (1996, pp. 235–237). in the light of what remains of Brüning’s papers indicates several others. Her suggestion (p. 237) that the Chase Bank may have been interested, despite some noises to the contrary, is probably correct. The Chancellor’s daily calendars (Tageszettel), now in the Brüning Papers, Harvard University Archives, HUB (FP) 93.35, Box 1, show that Shepard Morgan met twice with the Chancellor around this time. He was a Chase Bank official, though the daily calendar mistakenly identifies the bank. That he may have been scouting loan possibilities seems clear; see, e.g. J. E. Crane’s memorandum to Fed Governor Harrison of Jan. 24, 1931, Federal Reserve Bank of New York Archives, C261, Germany – German Government 1916–1950. This recounts a lunch with Morgan; while Morgan was skeptical then, he was obviously closely following events and probably shared Crane’s conviction that “in view of the recent improvement in the bond market here” one could reasonably expect that “foreign loans or credits may be somewhat more popular in the near future.” It is a puzzle to us how Bachmann reconciles the fact that banks were hounding the German government to get its business with her pessimism about German chances of floating loans. 35. See the handwritten memorandum, probably by T. H. McKittrick, dated Feb. 14, 1931 in 6–14 of Carton 6 of the Thomas H. McKittrick Collection, Baker Library, Harvard Business School. Mendelssohn conducted a great deal of business from Amsterdam after World War I. See Kupferberg (1972, p. 247) for Mendelssohn specifically and Aalders and Wiebes (1996, pp. 38–40) for the broader context. As mentioned below, Lee, Higginson was then very close to Ivar Kreuger, who at the time occupied a significant position in nothern European finance; it would not be surprising if Kreuger assisted Lee, Higginson in mustering support in Scandanavia. It does appear that he supported the Belgians and others who were working to move the BIS, though any discussion takes us too far afield. 36. Handwritten memorandum, Thomas H. McKittrick Collection, Baker Library, Harvard Business School, Feb. 14, 1931. Bachmann (1996), who relies on the official archives, does not mention Schacht as a significant force in the rejection of this proposal. On p. 238, she quotes a Mendelssohn partner’s advice to a German official on dealing with Lee, Higginson, but is unaware of the real context, which only emerges from the McKittrick Collection. From the latter, it appears the negotiations about a special relationship between the bank and the German government went on rather longer than indicated in the sources Bachmann cites. 37. “American” is used here for clarity, because the best known twentieth century examples come from the U.S. In the nineteenth century, such panics were not uncommon in many European countries, including Germany. 41
4259 Ch01 4/12/02 9:42 am Page 42
42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
38. The data come from “Zwischenbilanzen Deutscher Banken,” Wirtschaftsdienst, 16, passim. These are the same data that appear in Die Bank. They are reproduced in Institut für Konjunkturforschung (1933). For a general account of German banking in this period, see Whale (1930). 39. These figures include deposits from other banks and omit deposits of longer maturity. 40. Other, smaller assets include cash, Giro accounts, Lombards, and securities. None of them changed in any important way during the first half of 1931. 41. It should be noted that the total time deposits on 30 May 1931 from the individual banks are 100 million RM less than the total in Table 5. 42. See Forbes (2000, p. 33) and Morton (1943, pp. 30–33; 270–271). 43. The collapse of the Nordwolle was a process rather than event. We discuss some of the major dates below. But see the discussions in Born (1967) and Fiedler (1997). 44. Reichsbank to Federal Reserve Bank of New York, May 13, 1931. New York Federal Reserve Bank Archives, C261, “Germany-Reichsbank 1931–1931.” In the original translated cable, the order of the two words in brackets are reversed. The translations were often clumsy, with many showing subsequent hand corrections, and the meaning here is clear. We would be the last to deny that most official German spokespersons, including Reichsbank officials, normally took quite a different line in public and sometimes in private. But this is hardly surprising, since they could not well blame the crisis on their own actions without making everything much worse and even possibly bringing down the government. Still it is interesting that a month later, as its gold reserves disappeared in the rip tide of withdrawals and German public hand wringing about the Austrian crisis reached a crescendo, the Reichbank could still privately advise the Fed that “First reason for withdrawals probably certain political nervousness abroad with concomitant uncertainty concerning reparations.” See Luther to Harrison, June 11, 1931. New York Federal Reserve Bank Archives. C261 Policy and Procedures Germany-Reichsbank. 45. Reichsamt (1933a, p. 78) reports monthly flows in and out of savings banks. Reichsamt (1933b, p. 101) presents bimonthly data for cooperative banks. In JanuaryFebruary, there were net inflows; March-April and May-June show very small outflows of sizes that rule out talk of a run. After July, the size of the outflows increases greatly. But that only makes our point. 46. See the discussion in Whale (1930, pp. 334–339). 47. The currency data are from Reichsamt (1932, p. 326). The deposits are demand deposits for all credit banks, as discussed above. See also the more aggregrate data, which reveal the same pattern, in James (1984). 48. Various versions of the “structural weakness” view exist; the account in James (1986) certainly contains elements of such a view. Goedde (2000) presents a strong argument that German bankers over-lent in the conviction that they were “too big to fail.” 49. For bank stocks and the general market, see our discussion, above, Note 24; Born observed that banks owned industrial shares, but he had a totally different point in mind. See Born (1967, p. 59). 50. The Economist, May 9, 1931, p. 15; the article summarized “The German Banking Position.” On p. 8 of the same issue, the magazine wrote that “in the current year the marked upward movement in the share and loan market has been favorable to the banks
4259 Ch01 4/12/02 9:42 am Page 43
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
43
and their investments are worth more than at the end of last year.” For data on bank stocks and stocks in general, see (Reichsamt, 1932, pp. 354–355). 51. The celebrated Treaty of Locarno that had ushered in the famous Franco-German rapprochement of the mid-twenties only guaranteed borders in western Europe. For Brüning’s response to Briand, see, e.g. Winkler (1993, pp. 382ff.) A particularly succinct summary is Feuchtwanger (1995, pp. 239): “It had been an objective of Hindenburg and Schleicher in appointing Brüning that such a shift should occur. Their aims were fully shared by the Chancellor, though here, as in his move toward presidential government, caution was his watchword.” 52. The literature here is enormous; see, e.g. Winkler (1993, pp. 382ff.) A particularly acute contemporary overview is Wheeler-Bennett (1935, pp. 4–5). This British author was then in close touch with Brüning and many senior German military leaders. See Wheeler-Bennett (1974). 53. For the pocket battleship’s effect on the military balance, see below; for the domestic political problems, see, e.g. Winkler (1993, pp. 400ff.). The issue resonated strongly with voters on the left; by supporting the ship, the SPD opened itself to attack by the KPD. Because large navies were expensive, arms races also posed real threats to budgets in many countries. In the Spring of 1931, Lamont and other bankers were optimistic about Germany in part because of their confidence in the prospects for additional arms control agreements. See, e.g. other parts of the letter of March 10, probably by Lamont, quoted earlier or the reference to increased “prospects for political security in Europe” in the Morgan partners’ letter of March 14, 1931, above. Put simply, arms races raised country credit risks. 54. The need for foreign capital comes up several times in Cabinet deliberations in this period, recorded in the cabinet deliberations recorded in Koops (1982). 55. For the Depression and French colonial policy, see Jacques Marseille (1984, especially Chapters 6–9) and Bouvier, Girault and Thobie (1986, pp. 235ff.) On the exposition, which was a huge success and, in contrast to most such ventures, actually turned a profit, see Bouvier, Girault and Thobie (1986, p. 230). In the early thirties the French army’s Marshal Lyautey was a leader in attempting to unite business groups and the military behind policies to build up the empire. For German complaints about French “imperialism,” see, e.g. Stegmann (1978, p. 217–218); for specific German firms, see the discussion below. For the protests occasioned by German export pressures in Britain, see, among many sources, Forbes (2000, pp. 30–31 and 46–47). Forbes’ work deals mostly with the period after the British abandonment of gold, which was certainly a watershed, but the pressures were evident before the sterling devaluation. See the excellent discussion in Cain and Hopkins (1993, pp. 72–75). 56. The literature on German foreign policy and “Mitteleuropa” is too huge and contentious for this paper to attempt to review. Many studies, such as Kaiser (1980) have been dubious about the role economic factors played in these developments. But see Berghahn (1996, especially the Introduction) and Stegmann, “Mitteleuropa.” The latter’s discussion of splits within the business community, particularly over tariffs, and how France figured in the thinking of various groups bears careful reading. For German General Electric, the chemical industry, etc., see pp. 208ff. As Stegmann notes (p. 218), the famous German Mitteleuropäischen Wirtschaftstag, which agitated for German designs to the east, had been subsidized for a while by the German Foreign Office and supported by Schacht while President of the Reichsbank. It was reorganized in 1931 with financial support from Siemens, I. G. Farben, leading iron and steel firms, and a 43
4259 Ch01 4/12/02 9:42 am Page 44
44 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
major mining association. See below for additional references and discussion on the role of big business in these movements. The early German proposals definitely interested some French groups, though they imagined themselves in the leading role. For the French side, see, besides Stegmann, Knipping (1987, pp. 168–176), though, like Winkler (whose own treatment contains additional evidence on blocs within German business), we have some reservations about certain of his points. For I. G. Farben and the Brüning cabinet, see, e.g. Abraham (1986, p. 157), the sources cited earlier, but particularly the speech by Carl Duisburg discussed below, in support of the customs initiative. The vast buildup of gold reserves since the stabilization of the Franc gave the French state new abilities to project its financial power into central Europe. There is no doubt this worried German policy makers and business groups, though the French strategy of “encircling” Germany to the east was not always pursued with single-minded resolution. (See on this point, Wandycz (1988, pp. 178ff.) Still, some British policy makers who were favorably inclined toward Germany, were highly critical of the French in the run up to the summer 1931 crisis. See, e.g. Montagu Norman’s comments on French “encirclement” policy recorded in Henry Stimson’s Diary for April 8, 1931. The original of the diary is at Yale University Library; our reference is to the microfilm version, Reel 3. 57. Among the immense literature, see, e.g. Bennett (1962); Stambrook (1961); and Kaiser (1980). Stegmann, “”Mitteleuropa,” brings out the institutionalized cooperation between big business and the government, as well as the secrecy which enveloped their deliberations. For Brüning on the timing and the (false) claim that Germany could not choose it, see Stambrook (1961, p. 38). The cabinet deliberations are recorded in Koops (1982). 58. New York Times, March 26, 1931, p. 12. The remarkable speech of March 24, 1931 by Carl Duisberg of I. G. Farben defending the notion of a customs union with Austria (and other countries in southeastern Europe) placed it firmly in the context of other countries’ efforts to form tighter trade blocs and bluntly stated that reconciliation with France would come only after Germany moved east. Indeed, the speech spoke vaguely of an eventual bloc extending from Bordeaux to either Sofia or Odessa, depending on which version of the text one relies upon. See the (incomplete) text and notes reprinted in, e.g. Michalka and Niedhart (1992, pp. 245–246). Because this speech almost uncannily anticipated the subsequent direction of German policy, it has frequently been instanced as an example of the role big business played in German policy making. In recent years, however, a critical counter-current has emphasized the continuing attachment to international markets of some large firms, notably I. G. Farben and Siemens. See, e.g. Feldenkirchen (1999) for Siemens; and for I. G. Farben, Plumpe (1990, pp. 504–505) and Hayes (2001, p. 45). The latter acknowledges Stegmann’s (1978) observation that Farben helped to subsidize the Mitteleuropäischer Wirtschaftstag, but disputes its significance. For our paper, this discussion is not material, once one recognizes that the German government did attempt a customs union and that this step seriously damaged its relations with France. But Duisberg’s speech was no venture in cooperative internationalism. Hayes’ claim (p. 45) that Duisberg stressed the importance of an understanding with France misses the critical point that no one in the audience – or the French – could have: that while the Farben executive felt that an accord with France was important for a final (“endgültige”) resolution of European problems, he believed it could wait until after Germany, Austria, and southeastern Europe got moving on the “Zollunion” he advocated.
4259 Ch01 4/12/02 9:42 am Page 45
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
45
59. New York Times, March 26, 1931, p. 12. Compare a contemporary assessment in Dulles (1932, p. 239): “In fact, although March 22 has been called a turning point because of the Austro-German Customs proposal, the tragic consequences were not evident for about a month.” For the attempts at compromise, see Bennett (1962, pp. 96ff.). 60. For the discord over French attempts to buy into municipal utilities, see the ministerial discussion of March 27, 1931 in Koops (1982, Vol. II, pp. 994–996), which is quite clear about the embarrassing character of the problem. For the furor over the French Finance Minister’s remarks, see Freiherr von der Osten-Warnitz to President Hindenburg, March 2, 1931 in Opitz (1977, pp. 576–579). For the personal discussions between him and the President, see Meissner to Brüning, March 5, 2001 on p. 580; this indicates that the President directed that the letter be sent to the Chancellor, which was a characteristic way Hindenburg put pressure on him. See, e.g. Dorpalen (1964), who also notes that Hindenburg could often resist almost anything except pressures from his fellow Junkers. 61. See the discussion below on the Germans and the BIS; our dating follows the primary sources we cite; for the Kindersley Plan, see also Eichengreen and James (2001). 62. Gates W. McGarrah telegram sent through Morgan & Cie., to J. P. Morgan & Co., New York, March 18, 1931; copied to Harrison, Owen D. Young, and Jackson Reynolds. We have slightly regularized punctuation and spelling in the telegram. The telegram is in The Pierpont Morgan Library, New York, Morgan Bank European Papers, Box 13. 63. The first quotations in this paragraph come from the McGarrah’s telegram of March 18, just cited above. The American bankers response is from a telegram signed by J. P. Morgan, T. W. Lamont, and S. P. Gilbert; sent as a telegram from J. P. Morgan & Co. to Morgan et Cie., March 14, 1931; to be relayed back to McGarrah and confederates. The telegram also mentions that putting U.S. dollars in international institutions would undoubtedly arouse criticism. The telegram can be found in the The Pierpont Morgan Library, New York, Morgan Bank European Papers, Box 13. The Latin bankers’ objections are also recorded in McGarrah to J. P. Morgan & Co., March 18, 1931; also in Box 13. The Belgian efforts to move the French are related in Bussière (1992, pp. 397ff.). Details of the late Spring proposals to the BIS are in Carton B32221 of the French Ministry of Finance Archives, Savigny-le-Temple. These contain reports from the French consul at Basle. See, for example, the copy of his letter to the Minister of Foreign Affairs, April 21, 1931, which reports how the French blocked action on a proposal by Luther for BIS action in support of German export credits. A copy of his letter to the Minister of Foreign Affairs, May 20, 1931, reports on a BIS meeting in Basle in which a plan of the Belgian banker Franqui for more aggressive action by the BIS received support from Norman and Luther. The board nevertheless refused to act. See also the “Rapport du Comite de la BRI, tenu a Bruxelles, les 3 et 4 Juin, 1931” in the same Carton, in which, yet again, Melchior from Germany and Sir Charles Addis from the U.K. lined up behind Franqui, while the French opposed more vigorous intervention. 64. By itself, the customs union proposal rendered exports of long term capital impossible, but did not completely shut the market for short term capital. The nature of the German bind (before the budget blew up) was widely appreciated. As the Commercial and Financial Chronicle summed up the widely shared sentiment on the eve of the fall: “The chief difficulty of the Berlin money market is the disinclination of international bankers to make long-term issues.” See p. 3602 of the issue of May 16, 1931. 45
4259 Ch01 4/12/02 9:42 am Page 46
46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
65. German sources disagree slightly on the exact amount of the shortfall. Winkler (1993, pp. 403ff.) discusses Dietrich’s warning to Brüning. Bachmann (1996, pp. 249–250), cites internal documents indicating numbers close, but not identical, to those. The various ministerial discussions are recorded in Koops (1982). 66. “Besprechung im Reichstag vom 6. März, 1931,” Koops (1982, Vol. II, p. 928). See also the account by a top civil servant who was one of the very few people present, Hermann Pünder (1961, pp. 92–93). 67. Again, various German sources disagree very slightly on the exact numbers. Compare Bachmann (1996, p. 249) with James (1986, p. 73). The provision for further cuts in the event of budget shortfalls was noted earlier in the New York Times, March 26, 1931, p. 2, which at the time simply treated it as one more piece of good news. For Lee, Higginson’s reaction to the confidential news, see the telegram from the New York office to Courtney in Lee, Higginson’s Paris office, March 10, 1931, Thomas H. McKittrick Collection, Baker Library, Harvard Business School, Box 6: “Any move permitting interpretation that government was again borrowing would damage it greatly.” The reason is obvious: the world was just then celebrating Brüning’s budgetary triumph, with the key votes coming in the next days. 68. See “Private and Confidential,” [L,H Paris office telegram] to Murnane, March 10, 1931, Thomas H. McKittrick Collection, Baker Library, Harvard Business School: “Have therefore thought of approaching Moret . . .. As he himself has already stated he is much interested in a financial rapprochement between the two countries and naturally the whole world particularly perhaps France is interested in doing what is possible and practical to make as easy and comfortable as possible the internal money market in Germany.” “We could give him in confidence if possible some data about German requirements over next few months and the way German treasury bills are created by Reichstag approval and so forth . . ..” 69. “Billions” is, again, meant here in the American sense. The German text speaks of “Milliarden.” See, “Ministerbesprechung vom 29. May 1931,” Koops (1982, II, p. 1131). 70. On Brüning’s awareness that the American election would affect the timing of his plans, see “Besprechung vom 7. Mai, 1931,” Koops (1982, Vol. II, p. 1054; for the BIS, see p. 1056. The role of the American elections was not a secret; see Commercial and Financial Chronicle, May 16, 1931, p. 3636. Brüning and his advisors repeatedly discussed the BIS. See, e.g. the discussions of March 6, Koops (1982, Vol. II, p. 927) or June 3 (p. 1179). 71. “Besprechung vom 7. Mai, 1931,” Koops (1982, Vol. II, p. 1054). In the protocol, what Brüning actually says is is “dann aber werde sich keine Regierung der Mitte finden, die bereit sei, dem Rechtskabinett die Verantwortung abzunehmen.” The Labor Minister, who was close to Brüning, agreed completely on the need to move on reparations. See p. 1055. 72. “Besprechung vom 7. Mai, 1931,” Koops (1982, Vol. II, pp. 1058–1059). 73. For the diplomatic issues, see especially, Bennett (1962). For the earlier stages of Brüning’s decision, see Koops (1982, Vol. II, pp. 925–928). For the connection between the British talks and the domestic crisis, see Winkler (1993, pp. 404ff.). 74. James (1986, p. 306) makes this claim. But the April tax collection numbers had been published in May and were recognized at once as a serious problem. See the discussion and references below. 75. See, e.g. The Economist, May 9, 1931, p. 1001. The influence of particular situations should not be exaggerated, however. As noted above, German savings banks
4259 Ch01 4/12/02 9:42 am Page 47
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
47
actually held more deposits at the end of May than they did at the beginning of the month. 76. They were published regularly and quickly by several publications, including Wirtschaft und Statistik and often in daily papers, such as the Frankfurter Zeitung. Major foreign papers carried also discussed the dismal numbers. See, e.g. the New York Times report from Berlin on May 15 as excerpted in the Commercial and Financial Chronicle, May 23, 1931, p. 3810: “German Tax Revenue Below Estimate – Shortage of 500,000,000 Marks Now Indicated – Expenditure To Be Cut.” In most recent literature by economic historians, the proposal for a customs union receives at most, passing attention. It is not integrated into a broader analysis, so that its pivotal significance emerges. It is worth noting, however, that some, but not all, contemporary accounts did realize the significance of the customs union proposal. Besides the New York Times story quoted above, see the discussion in Einzig (1935, p. 217), though he also touts the role of the Creditanstalt. Much sharper formulations can be found in Moulton and Pasvolsky (1932, p. 308) and especially Salter (1932). The latter, an influential figure in British financial and political circles, wrote (p. 226) that “in June, as the conditions tending to a crisis developed, men were asking each other in New York, as they had been six months before in Paris, whether there was a danger of a new war in Europe. The addition of these fears to the existing doubts as to the solvency of debtors impoverished by the depression gave a fatal shock to confidence.” 77. See, e.g. the following articles on the budget crisis in the Frankfurter Zeitung: “Wie Wird das Defizit des Reichsetats gedeckt werden,” on p. 1 of the May 6 issue; “Bemerkungen, Das Defizit,” on p. 1, May 21, 1931, along with more detailed material on p. 3. This contains a blunt assessment of the meaning of the April revenue shortfall. 78. See The Economist, May 2, 1931, p. 944; for the failure, see The Economist, May 23, 1931, p. 1102. This is vague about the date, but not the flop. The issue came out on May 4, 1931; see the data in Rosenstiel (1931b, p. 131, Table). 79. For the Prussian SPD, the street clashes, and the Minister of the Interior, see Winkler (1993, pp. 403–405). 80. Compare Brüning’s statement to his ministers on May 7 with the tone of the meeting on June 3, particularly the Finance Minister’s comments, which neither Brüning nor anyone else bothers to rebut. The latter meeting also witnesses much grasping at straws, even a mention of the possibility that Norman would finally persuade the BIS to move seriously. 81. For examples where Brüning flatly denied he planned to change policy, see, e.g. his May 7 assertion that a change in policy would be fatal, Koops (1982, Vol. II, pp. 1056–1057) or his very plain spoken June 2 comments (p. 1164). For the Finance Minister, see, e.g. pp. 1178–1179; especially N. 1 on p. 1179 for the threat to resign. For contrasting public statements by him and the Chancellor, compare Brüning in the New York Times, May 11, 1931, p. 10, with Dietrich’s remarks on the inevitability of reparations concessions, Commercial and Financial Chronicle, May 16, 1931, p. 3636. 82. For the rallies, see, e.g. Feuchtwanger (1995, p. 251). Given the Versailles Treaty’s restrictions on the size of the German army, the Army believed it had to have the cooperation of the Stahlhelm in border regions (p. 234) and, as shortly became apparent, it was also becoming far more interested in cooperation with the Nazi SA. Some prominent industrialists were members of the Stahlhelm; Hindenburg was long its nominal head, though the Stahlhelm leadership often considered him too moderate. For 47
4259 Ch01 4/12/02 9:42 am Page 48
48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
the French election, see, e.g. Winkler (1993, p. 406). On systemic risk, see especially Salter’s remarks, quoted above. 83. For the effect on French naval planning, see Toynbee (1932, p. 71 ). For the collapse of the French-Italian naval agreement in the wake of the proposal for a customs union, see pp. 273–274. The pocket battleship created special problems because it was said either to be able to sink or outrun any ship afloat. See Wheeler-Bennett (1932, pp. 336–337). This was a major reason why Brüning’s decision, mentioned above, to press ahead with another was so controversial. It threatened a multinational round of big new arms spending. 84. For Gröner’s warning, and the urgency of heeding it, see Bennett (1962, pp. 120–121). For Voegler’s speech and the meeting, see the account in the London Times, June 4, 1931, p. 13. Other prominent industrialists, including Thyssen and Reusch, who were present, criticized Brüning for being weak and called for authoritarian methods and an anti-socialist drive. The article also discusses Hjalmar Schacht’s promotion of a moratorium before a group of industrial leaders. 85. For the dates, see Toynbee (1932). 86. A “Note du Mouvement des Fonds” of the 26th of May, 1931, bearing the title “Reparations et Dettes Interalliees,” now in the archives of the Banque de France (in a series of cartons numbered “59)” shows that the French had taken the hints in the German press and expected that Brüning would raise the reparations issue with the British. For the fall in Young Plan bonds, see Toynbee (1932, p. 68); this also mentions the Stahlhelm rally discussed earlier. 87. See, e.g. the summary in Woodward and Butler (1947, Vol. II, p. 70). 88. The London Times account of the decree also reported the official denial. See the Times, June 6, 1931, p. 11. The Commercial and Financial Chronicle, June 13, 1931, p. 4293, quoted a story by the Associated Press forecasting an imminent move on reparations by Germany, despite the denials. The Chronicle, p. 4293, also noted that the manifesto for the first time referred to reparations as “tribute.” 89. The deal was between the government and a syndicate led by Ivar Kreuger, the “Match King.” Kreuger and Lee, Higginson were so closely linked that the subsequent bankruptcy of the one spelled the ruin of the other. The $75 million dollar payment was not a surprise; but the New York Times announcement came on May 28, 1931, as the paper also reported on the previous day’s sales of German assets. See p. 43 of that issue. Simple arithmetic and a comparison with the other entries for Reichsbank total gold exchange reserves for May in our Table 8 confirms the Reichsbank’s statement to the New York Federal Reserve Bank quoted earlier on the insignificance of German bank relations with the Creditanstalt and rules out traditional stories of chain bankruptcy spreading from Austria to Germany. Note that not later than about May 25, both German newspapers and French archives agree that word of Brüning’s shift in course began leaking. As observed above, this clearly led to a first, and by later standards, rather modest, wave of selling of Mark assets in world financial markets, which does not require invoking the Creditanstalt. It is also worth noting that in a telephone conversation with Harrison of the New York Fed on June 20, 1931, Montagu Norman indicated that end of the month demands on the Reichsbank were normally heavier; the remark appears to be meant generally and not to be simply about the June situation. See New York Federal Reserve Bank Archives, File 3115.2; this is a file of memoranda by Harrison on conversations with Norman. Late Spring reverses in the American market may have made some U.S. investors nervous and led them to bring some money home. Federal Reserve behavior in this
4259 Ch01 4/12/02 9:42 am Page 49
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
49
period does suggest some anxiety and was markedly at variance with its actions during the rest of the year. Cf. the remark on this period in Epstein and Ferguson (1984). But Table 8 and our earlier discussion of banking data also rule out significant influences on the German situation from this quarter in May. 90. The characterization comes from the Berliner Tageblatt, cited in Mommsen (1996, p. 387. 91. For the possibility of running out of money, the fears of Groener and others of a revolution, the possibility of military intervention, etc., see Winkler (1993, pp. 408–413) and Feuchtwanger (1995, pp. 246–248). For Schleicher’s meeting with Brüning, see the Chancellor’s official calendar for June 13, 1931, Brüning Papers, Harvard University Archives, HUG (FP) 93.35, Box 1 of 6. A note there explains that the Reichswehr was calling in the commanders of the various military districts for consultations on procedures in case of civil unrest. 92. See, again, Winkler (1993, pp. 408–413) and Feuchtwanger (1995, pp. 246–248). 93. Temin (1989); see also Born (1967, pp. 74ff.) For the gold and foreign exchange cover, see our Table 8 and Reichsamt (1932, p. 326). Born quotes slightly different figures that do not affect any issue under discussion; see his discussion on p. 74. The actual cover requirement was somewhat complex. Reichsbank notes in circulation required a minimum cover of 40% in gold or foreign exchange. Foreign exchange could not comprise more than 25% of the total cover; this is equivalent to a 30% gold cover rule. See Madden and Nadler (1935, pp. 373ff.) which also discusses the limited role played by the handful of banks that still had powers to issue notes. The entry in our Table 4 for Young Bonds is somewhat misleading, in that it reflects the momentary revival brought about toward the end of the month by the initial offer of the Hoover Moratorium. In mid-June, the bonds fell as low as 65. Later in the summer, they collapsed, of course. By September, they languished in the thirties. 94. Balderston argued similarly that the problem with the German banks was the Reichsbank’s inability to supply reserves, but he did not connect the Reichsbank’s position in June with Weimar’s political act of March or to the German budgetary crisis. See Balderston (1993) and Balderston (1994). 95. Fleming (2000, p. 147). Fleming’s article details his notes on Keynes’ lectures in Geneva during July, 1929. 96. Note that both Brüning and Luther mention the possibility of revaluing gold in the meetings just prior to the fatal Notverordung. See Koops (1982, Vol. II, p.1058). It appears they were looking for some kind of international agreement on this, but the discussion is too abbreviated to be sure. It is perhaps of interest that Brüning’s appointment books indicate that Sir Walter Layton was the last official visitor he saw before the American ambassador arrived to discuss the text of the famous German appeal for help to Hoover and the first person the Chancellor saw afterward. 97. See Borchardt (1991); for the critics, see especially, Holtfrerich (1990) and Kim (1997).
36 37
ACKNOWLEDGMENT
38 39 40
We are grateful to seminar participants at Oxford, the London School of Economics, Harvard, and the Wharton School for helpful comments. 49
4259 Ch01 4/12/02 9:42 am Page 50
50 1
THOMAS FERGUSON AND PETER TEMIN
REFERENCES
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
Aalders, G., & Wiebes, C. (1996). The Art of Cloaking Ownership. Amsterdam: University of Amsterdam Press. Abraham, D. (1986). The Collapse of the Weimar Republic (2nd ed.). New York: Holmes & Meier. Bachmann, U. (1996). Reichskasse und öffentlicher Kredit in der Weimarer Republik: 1924–1932. Frankfurt am Main: Peter Lang. Balderston, T. (1993). The Origins and Course of the German Economic Crisis: November 1923May 1932. Berlin: Halde & Spener. Balderston, T. (1994). The Banks and the Gold Standard in the German Financial Crisis of 1931. Financial History Review, 1, 43–68. Bennett, E. (1962). Germany and the Diplomacy of Financial Crisis. Cambridge, MA: Harvard University Press. Berghahn, V. (1996). Quest for Economic Empire: European Strategies of German Big Business in the Twentieth Century. Providence, Rhode Island: Berghahn Books. Borchardt, K. (1991). Perspectives on Modern German Economic History and Policy. Translated by Peter Lambert. Chicago: University of Chicago Press. Born, K. (1967). Die deutsche Bankenkrise 1931: Finanzen und Politik. Munich: R. Piper. Bouvier, J., Girault, R., & Thobie, J. (1986). L’impérialisme à la française. Paris: Editions la decouverte. Bratz, M. (1958). Der deutsche Privatbankierstand in der Nachkriegzeit: 1918–1933. Berlin: Paul Brandel. Burke, B. (1994). Ambassador Frederic Sackett and the Collapse of the Weimar Republic, 1930–1933. New York: Cambridge University Press. Bussière, É. (1992). La France, la Belgique et l’organisation économique de l’Europe: 1918–1935. Paris: Ministère de l’économie des finances et du budget: Comité pour l’histoire économique et financiere de la France. Cain, P. J., & Hopkins, A. G. (1993). British Imperialism: Crisis and Deconstruction: 1914–1990. New York: Longman. Deutsche Bundesbank (1976). Währung und Wirtschaft in Deutschland: 1876–1975. Frankfurt am Main. Deutsches Institute für Wirtschaftsforschung (1932). Wirtschaftszahlen 1925 bis 1931. Berlin: R. Hobbing. Dorpalen, A. (1964). Hindenburg and the Weimar Republic. Princeton: Princeton University Press. Dulles, E. (1932). The Bank for International Settlements at Work. New York: Macmillan. Eichengreen, B. (1992). Golden Fetters: The Gold Standard and the Great Depression. New York: Oxford University Press. Eichengreen, B., & Hatton, T. (Eds) (1988). Interwar Unemployment in International Perspective. Dordrecht: Kluwer. Eichengreen, B., & James, H. (2001). International Monetary and Financial Reform in Two Ages of Globalization. In: J. Williamson & J. Taylor (Eds), Globalization and History. Chicago: University of Chicago Press. Einzig, P. (1935). World Finance: 1914–1935. New York: Macmillan. Ellis, H. (1941). Exchange Control in Central Europe. Cambridge: Harvard University Press. Epstein, G., & Ferguson, T. (1984). Monetary Policy, Loan Liquidation, and Industrial Conflict: The Federal Reserve and the Open Market Operations of 1932. Journal of Economic History, 51, 190–200.
4259 Ch01 4/12/02 9:42 am Page 51
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
51
Feldenkirchen, W. (1999). Siemens, 1918–1945. Columbus, Ohio: Ohio State University Press. Feinstein, C., & Watson, K. (1995). Private International Capital Flows in Europe in the Inter-War Period. In: C. Feinstein (Ed.), Banking, Currency, and Finance in Europe Between the Wars (pp. 94–130). Oxford: Clarendon Press. Ferguson, T. (1995). Golden Rule: The Investment Theory of Party Competition and the Logic of Money-Driven Political Systems. Chicago: University of Chicago Press. Feuchtwanger, E. (1995). From Weimar to Hitler Germany, 1918–1933 (2nd ed.). New York: St. Martin’s. Fiedler, M. (1997). Business Scandals in the Weimar Republic. In: W. Feldenkirchen & T. Gourvish (Eds), European Yearbook of Business History (pp. 155–178). Flechtner, H.-J. (1959). Carl Duisberg: vom Chemiker zum Wirtschaftsführer. Düsseldorf: Econ Verlag. Fleming, G. (2000). Foreign Investment, Reparations, and the Proposal for an International Central Bank: Notes on the Lectures of J. M. Keynes in Geneva, July 1929. Cambridge Journal of Economics, 24, 139–151. Forbes, N. (2000). Doing Business with the Nazis: Britain’s Economic and Financial Relations with Germany: 1931–1939. London: Frank Cass. Goedde, I. (2000). The German Twin Crises of 1931. Manuscript. Goldstein, M., Kaminsky, G., & Reinhart, C. (2000). Assessing Financial Vulnerability: An Early Warning System for Emerging Markets. Washington, D.C.: Institute for International Economics. Grossman, R. (1994). The Shoe That Didn’t Drop: Explaining Banking Stability During the Great Depression. Journal of Economic History, 54, 654–682. Harris, C. (1935). Germany’s Foreign Indebtedness. London: Oxford University Press. Hayes, P. (2001). Industry and Ideology: I. G. Farben in the Nazi Era (2nd ed.). New York: Cambridge University Press. Heyde, P. (1998). Das Ende der Reparationen: Deutschland, Frankreich und der Youngplan: 1929–1932. Paderborn: F. Schoningh. Holtfrerich, C. (1990). Economic Policy Options and the End of the Weimar Republic. In: I. Kershaw (Ed.), Weimar: Why Did German Democracy Fail? New York: St. Martin’s. Institut für Konjunkturforschung (1933). Konjunktur-statistisches Handbuch. Berlin: Verlag von Reimar Hobbing. Institute of International Finance (1931). Credit Position of Germany. New York: Institute of International Finance. James, H. (1984). The Causes of the German Banking Crisis of 1931. Economic History Review, 38, 68–87. James, H. (1986). The German Slump Politics and Economics: 1924–1936. Oxford: Clarendon Press. James, H. (2001a). The Deutsche Bank and the Nazi Economic War against the Jews. Cambridge: Cambridge University Press. James, H. (2001b). The End of Globalization. Cambridge MA: Harvard University Press. Kaiser, D. (1980). Economic Diplomacy and the Origins of the Second World War. Princeton: Princeton University. Kim, H.-I. (1997). Industrie, Staat, und Wirtschaftspolitik: Die konjunkturpolitische Diskussion in der Endphase der Weimarer Republik. Berlin: Duncker & Humblot. Kindleberger, C. (1986). The World in Depression (2nd ed.). Berkeley: University of California Press. Knipping, F. (1987). Deutschland, Frankreich und das Ende der Locarno Ara, 1928–1931. Munich: Oldenbourg.
51
4259 Ch01 4/12/02 9:42 am Page 52
52 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
THOMAS FERGUSON AND PETER TEMIN
Kolb, E. (1988). The Weimar Republic, P. S. Falla (Trans.). London: Unwin Hyman. Kolb, E. (1993). Die Weimarer Republik (3rd ed.). Munich: R. Oldenbourg Verlag. Koops, T. (Ed.) (1982). Die Kabinette Brüning, Akten der Reichskanzlei (Vol. 2). Boppard am Rhein: H. Boldt Verlag. Kupferberg, H. (1972). The Mendelssohns. New York: Scribner’s. Länderrat des Amerikanischen Besatzungsgebiets (Ed.) (1949). Statistisches Handbuch von Deutschland 1928–1944 (Vol. II). Munich: Franz Ehrenwirth-Verlag. Madden, J., & Nadler, M. (1935). The International Money Markets. New York: Prentice Hall. Marseille, J. (1984). Empire colonial et capitalisme français. Paris: Albin Michel. McNeil, W. (1986). American Money and the Weimar Republic. New York: Columbia University Press. Michalka, W., & Niedhart, W. (1992). Deutsche Geschichte 1918–1933: Dokumente zur Innenund Aussenpolitik. Frankfurt am Main: Fischer Verlag. Mommsen, H. (1996). The Rise and Fall of Weimar Democracy, E. Forster & L. Jones (Trans.). Chapel Hill, N.C.: University of North Carolina Press. Morton, W. (1943). British Finance, 1930–1940. Madison, Wisconsin: University of Wisconsin Press. Moulton, H., & Pasvolsky, L. (1932). War Debts and World Prosperity. New York: Century Co. for the Brookings Institution. Opitz, R. (Ed.) (1977). Europastrategien des deutschen Kapitals: 1900–1945. Köln: Pahl-Rugenstein Verlag. Plumpe, G. (1990). Die I. G. Farbenindustrie AG: Wirtschaft, Technik und Politik: 1904–1945. Berlin: Duncker & Humblot. Puender, H. (1961). Politik in der Reichskanzlei. Stuttgart: Deutsche-Verlags Anstalt. Reichsamt, Statistisches (1932). Statistisches Jahrbuch für das deutsche Reich. Berlin: R. Hobbing Reichsamt, Statistisches (1933a). Die deutschen Sparkassen in den Jahren 1931 und 1932. Vierteljahrshefte zur Statistik des deutschen Reichs, 42(2), 76–100. Reichsamt, Statistisches (1933b). Die gewerblichen Genossenschaftsbanken in den Jahren: 1931/1932. Vierteljahrshefte zur Statistik des deutschen Reichs, 42(2), 101–105. Reichsamt, Statistisches (1933c). Deutsche Wirtschaftskunde: Ein Abriss der deutschen Reichsstatistik (2nd ed.). Berlin: R. Hobbing. Ritschl, A. (1998). Reparations Transfers, the Borchardt Hypothesis, and the Great Depression in Germany. European Review of Economic History, 2, 49–72. Rosenstiel, F. (1931a). Die Kreditmärkte. Wirtschaftskurve, 10 (April), 17–21. Rosenstiel, F. (1931b). Die Kreditmärkte. Wirtschaftskurve, 10 (July), 125–131. Salter, A., Sir (1932). Recovery the Second Effort. London: G. Bell and Sons. Schmidt, C. (1934). German Business Cycles, 1924–1933. New York: National Bureau of Economic Research. Schuker, S. (1988). American ‘Reparations’ to Germany, 1919–1933: Implications for the Third World Debt Crisis. Princeton: Princeton University Department of Economics: Princeton Studies in International Finance. Southern, D. (1981). The Impact of the Inflation: Inflation, the Courts, and Revaluation. In: R. Bessel & E. J. Feuchtwanger (Eds), Social Change and Political Development in Weimar Germany. London: Croom Helm. Stachura, P. (Ed.) (1986). Unemployment and the Great Depression in Weimar Germany. New York: St. Martin’s. Stambrook, F. (1961). The German-Austrian Customs Union Project of 1931: A Study of German Methods and Aims. Journal of Central European History, 21, 15–44.
4259 Ch01 4/12/02 9:42 am Page 53
German Currency Crisis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
53
Stegmann, D. (1978). Mitteleuropa: 1925–1934. In: D. Stegmann, B.-J. Wendt & P.-C. Witt (Eds), Industrielle Gesellschaft und politisches System: Festschrift für Fritz Fischer (pp. 203–221). Bonn: Verlag Neue Gesellschaft. Temin, P. (1989). Lessons from the Great Depression. Cambridge MA: MIT Press. Thomas, B. (1934). Germany. In: H. Dalton, B. Thomas, J. Reedman, J. Hughes & W. Leaning (Eds), Unbalanced Budgets. London: George Routledge. Toynbee, A. (Ed.) (1932). Survey of International Affairs, 1931. London: Oxford University Press. Treue, W. (1972). Das Bankhaus Mendelssohn als Beispiel einer Privatbank im 19. und 20. Jahrhundert. In: Mendelssohn-Gesellschaft (Ed.), Mendelssohn Studien (pp. 29–80). Berlin: Duncker & Humblot. U.S. Department of State (1945). Papers Relating to the Foreign Relations of the United States 1930 (Vol. III). Washington, D.C.: U.S. Government Printing Office. U.S. Department of State (1986). Microfilm. Confidential U.S. State Department Central Files: Germany, Internal Affairs, 1930–1941. University Publications of America. Wandycz, P. (1988). The Twilight of French Eastern Alliances: 1926–1936. Princeton: Princeton University Press. Whale, P. (1930). Joint Stock Banking in Germany. London: Frank Cass. Reprint, Augustus Kelley. Wheeler-Bennett, J. (1932). Disarmament and Security since Locarno: 1925–1931. New York: Macmillan. Wheeler-Bennett, J. (1935). The Pipe Dream of Peace. New York: William Morrow. Wheeler-Bennett, J. (1974). Knaves, Fools, and Heroes. London: Macmillan. Winkler, H. A. (1993). Weimar 1918–1933: Die Geschichte der ersten deutschen Demokratie. Munich: Verlag C. H. Beck. Woodward, E., & Butler, R. (Eds) (1947). Documents on British Foreign Policy: 1919–1939, Second Series, Vol. 2. London: His Majesty’s Stationery Office.
22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
53
4259 Ch02 4/12/02 9:43 am Page 55
1 2 3 4 5 6 7 8 9 10
THE HOUSEHOLD BALANCE SHEET, CREDIT, AND UNCERTAINTY AT THE ONSET OF THE GREAT DEPRESSION IN THE USA
11 12 13 14
David Greasley and Jakob B. Madsen
15 16 17 18
ABSTRACT
19 20 21 22 23 24 25 26 27
The influence of the household balance sheet, the supply of credit, and uncertainty on consumer spending during the early years of the Great Depression in the USA are assessed within a unified life-cycle consumption function framework. Income uncertainty played the dominant role in the spending declines of 1930 and 1932. The depletion of households’ financial assets contributed modestly to the consumption falls in 1931–1932. Indebtedness, the severe penalties surrounding installment debt default, and the supply of credit had little effect on the consumer spending slump.
28 29
I. INTRODUCTION
30 31 32 33 34 35
A key strand in the historiography of the Great Depression highlights the role played by American consumers (Eichengreen, 1992, pp. 215–216; Fearon, 1987, pp. 151–153). Their real spending fell precipitously, by 19.3% in the period 1929–1933 (Lebergott, 1996, pp. 148–159). Considerable effort has been directed towards understanding this consumption fall since Temin (1976, pp. 62–83)
36 37 38 39 40
Research in Economic History, Volume 21, pages 55–77. Copyright © 2003 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623-0993-8
55
4259 Ch02 4/12/02 9:43 am Page 56
56 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
DAVID GREASLEY AND JAKOB B. MADSEN
emphasized its central role in the Great Depression, and failed to explain its causes. Subsequent historiography has debated whether uncertainty surrounding households’ incomes, consumers’ financial liquidity, credit supplies, or the severity of debt default penalties were causes of the collapse, but, typically, the effects are evaluated individually. This article assesses possible explanations of the consumption decline by considering the competing hypotheses within unified models. The approach estimates life-cycle-type consumption functions for various categories of spending, including durables, semi-durables, perishables, and services. Incorporating liquidity and uncertainty measures within the same consumption function allows their respective contributions to the decline of consumption during the period 1930–1933 to be gauged. Romer (1990), Olney (1999) and Mishkin (1978) provide different interpretations of the consumption collapse during the early years of the Great Depression. Mishkin’s (1976) liquidity hypothesis, potentially, has roles for both the risks associated with illiquidity, and income uncertainty, in explaining the falls in durable purchases. Mishkin’s (1978) model postulates that households’ risks of distress selling of their durable possessions are related to their holdings of financial assets, their debts, but also to their expected income and its variance. He shows that consumers are less likely to purchase durables the higher their debts, the lower their financial asset holdings, and the greater the uncertainty surrounding their expected income. Elsewhere, and rather differently, Mishkin (1990) contends that financial crises, by promoting information asymmetries and adverse selection, played an important role in prolonging the contraction through 1932.1 His measure, based on the spread of interest rates paid by high and low quality borrowers, indicates that lending risks of financial institutions rose sharply in 1931–1932, implying that financial disintermediation may have accentuated the collapse in spending. While much of this discussion centers on the implications for investment, Mishkin (1990, p. 23) notes the relevance of his argument for Romer’s (1990) interpretation of the consumer durables spending. Bernanke (1983a) also highlights the possible importance of financial disintermediation for explaining the great contraction, and utilizes alternative indicators, including the deposits of suspended banks. Romer (1990) takes a different view, and her influential discussion of income uncertainty drew on Bernanke’s (1983b) concept of irreversibility to explain why consumers reduced their durable purchases in 1930. The theory of irreversible purchases considers the value of waiting relative to the immediate acquisition of durables, and shows how heightened uncertainty may lead to the deferring of purchases of durable consumption goods. For Romer, the rise in share price volatility in the year following the Wall Street Crash indicates the
4259 Ch02 4/12/02 9:43 am Page 57
Household Balance Sheet in U.S. Depression 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
57
income uncertainty that led to the decline in consumer durable spending in 1930. Both Romer’s approach invoking the theory of irreversible purchases, and Mishkin’s liquidity model, postulate that income uncertainty reduced durable goods spending. Their positions are separated by Mishkin’s empirical finding that liquidity effects explain most of the durable spending fall during the first year of the Great Depression, and his denial that there was a sharp rise in uncertainty in 1930. He explains around 69% of reduced consumer durable and residential housing spending in 1930 by liquidity balance-sheet effects. Romer does not deny completely that liquidity effects reduced consumption in 1930. Her model though does not allow explicitly for liquidity effects, and shows that heightened income uncertainty explains fully the fall in durable spending in 1930. For Mishkin, uncertainty played a more important role in later years, following the banking and gold standard crises of 1931. Ferderer and Zalewski (1994) agree that the financial crises heightened uncertainty. They show that interest rate volatility was unimportant in 1930 but rose sharply in 1932–1933 as the doubts surrounding the USA’s adherence to the gold standard strengthened. Whether uncertainty played an important role in reducing consumption in 1932 deserves further attention. This year witnessed the largest consumption falls of the Great Depression, and Greasley, Madsen and Oxley (2001) show that both life cycle and permanent income consumption functions, for both durable and non-durable spending, have large negative residuals in 1932.2 New research by Lebergott (1996) shifts the attention from durables by emphasizing the breadth of the post-1929 decline across most categories of consumer spending. Olney (1999) extends the liquidity model to non-durables by postulating that consumers might reduce their propensity to consume, and thus their non-durable consumption, to avoid default on installment debt. Noting the unprecedented rise of installment debt to 1929, she argues that severe debt default penalties led consumers to curtail all spending in 1930 to reduce the likelihood that their durable goods would be repossessed. Ostensibly, Olney’s results explain most of Temin’s autonomous fall in 1930 consumption, although her empirical model excludes uncertainty. Brown (1997) reinforces her view that households were concerned greatly by their debts in 1930, postulating that there was a sharp rise in the voluntary early repayment of debt in that year. A fear of default, though, might not be the only reason why consumers reduce their non-durable consumption. Hahm and Steigerwald’s (1999) theoretical model and empirical findings show that consumers adjust their precautionary savings and non-durable consumption in response to changes in income uncertainty. A variety of explanations have been put forward to explain the consumption slump at the onset of the Great Depression. Here we estimate the effects of 57
4259 Ch02 4/12/02 9:43 am Page 58
58 1 2 3 4 5
DAVID GREASLEY AND JAKOB B. MADSEN
heightened uncertainty and deteriorating household balance sheets within a unified life-cycle consumption function framework using Lebergott’s new data for consumers’ expenditures. Romer’s share price volatility measure of income uncertainty is used, and new measures of financial assets and liabilities are estimated and deployed.
6 7 8
II. UNIFIED MODELS OF CONSUMPTION
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
Most studies of the consumption decline at the onset of the Great Depression, including those of Temin (1976), Mayer (1978) and Olney (1999), are based on the life-cycle hypothesis (LCH) of Ando and Modigliani (1963). The model used here adapts the life-cycle hypothesis to consider explicitly the ratio of consumption to income within a modern time series framework. While the aggregate consumption-income ratio tends to be stationary in the long run, counter-cyclical variations in the ratio are an implication of the LCH, because a cyclical decline in income is expected to be temporary, and the ratio of consumption and income will consequently increase in recessions. The important issue is to gauge whether heightened uncertainty, illiquidity, or credit constraints at the onset of the Great Depression disrupted the consumption-income ratio, to result in a catastrophic fall of consumption that was largely independent of changes in income and wealth. Much of the historiography of the consumption collapse investigates particular components of total consumption. The ratios of durable and various categories of non-durable consumption to income may vary in the long run, and the model adopted allows for this possibility. The model for categories of non-durable consumption is formulated as follows:
28 29
ln Ct = 0 + 1 ln Yt + 2 ln FASt + 3 ln LIAt + 4Vt
30 31
+ 5 (ln Ct – 1 – ln Yt – 1) + 6 (ln Ct – 1 – ln At – 1) +
(1)
32 33 34 35 36 37 38 39 40
where C is real per capita consumption, Y is per capita real disposable income, FAS and LIA are households’ per capita financial assets and liabilities, V is uncertainty, A is real per capita net value of household assets and is a disturbance term. This model allows for the dynamic adjustment of consumption to changes in income and wealth, with the error correction terms (ln Ct1ln Yt1), and (ln Ct1ln At1) certifying that consumption gravitates towards its equilibrium relationship with income and net wealth in the long run. A decrease in income, which is anticipated to be temporary, may lead to a temporary rise
4259 Ch02 4/12/02 9:43 am Page 59
Household Balance Sheet in U.S. Depression 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
59
in the C/Y ratio; however, the error-correction term certifies that a constant C/Y is maintained in the long run. For consumption of durables and semi-durables, their prices relative to that of total consumption are added to Eq. (1) and the real interest rate is added to the equation for durables. Hamburger (1967) derived the desired stock of consumer durables as a function of income, wealth, relative prices, and the real rate of interest. Using a simple stock adjustment mechanism he shows durables’ consumption to depend on these variables, and the stock of durables in the previous period. Thus Eq. (1), extended to durables and semi-durables, includes relative prices, and the real interest rate. In addition, in the case of durables, bank suspensions are included to investigate the possibility that financial disintermediation influenced this category of spending.3 In sum, the formulation of the LCH for non-durables specified as Eq. (1), investigates, how, if at all, influences other than income collapse contributed to the consumption decline at the onset of the Great Depression. Including uncertainty and liquidity variables in a life-cycle model provides a natural route to investigating the competing hypotheses of the consumption downturn at the onset of the Great Depression. For the balance sheet variables, A, the real per capita net value of household assets, is decomposed into financial assets and liabilities for the estimation of the empirical models since Mishkin (1978, p. 927) argues that a rise in debt has more powerful effects on consumption than a fall in financial assets. In the case of durables Romer’s uncertainty hypothesis considers whether income uncertainty reduced spending by increasing the value of waiting relative to spending immediately on irreversible purchases. However, the uncertainty hypothesis can be easily extended to non-durables. Hahm and Steigerwald show that an increase in income uncertainty lowers the consumption of non-durables under the assumption that consumers suffer from absolute risk aversion. Thus, a rise in precautionary savings in circumstances of heightened income uncertainty may lead to falls in all types on consumer spending.4 Rather differently, Mishkin (1990) postulates that financial institutions’ uncertainty surrounding their risks led to adverse selection and reduced lending, including to consumers for the purchase of durables. He uses the interest rate spread between a Treasury and a Moody’s Baa corporate bond to indicate information asymmetries between borrowers and lenders. However, the interest rate spread measures the uncertainty associated with fixed investment, default risk, and other factors, and is, therefore, not directly related to the adverse selection and financial disintermediation.5 Accordingly, Bernanke’s (1983a) measure of credit barriers, the deposits of suspended banks, is used in the following empirical sections. 59
4259 Ch02 4/12/02 9:43 am Page 60
60 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
DAVID GREASLEY AND JAKOB B. MADSEN
Olney’s extension of the liquidity hypothesis to non-durables postulates that installment debt had greater effects on their consumption before 1932 because default penalties were severe during these years. In the case of non-durables the simple uncertainty hypothesis maintains that consumers increase their precautionary savings by reducing non-durable consumption in circumstances of heightened income uncertainty. Olney, alternatively, argues that indebtedness has time varying effects depending on the severity of default penalties on installment credit. When penalties were severe, Olney argues consumers will reduce all spending, including that on non-durables, to avoid uncompensated repossession of their durable assets. Default penalties gradually became more severe through the 1930s, but Olney’s results show it was 1938 before installment debt positively influenced consumption at conventional levels of statistical significance. Her findings also show a coefficient for installment debt for the years to 1932 that is statistically significantly different to that for post-1938, although individually the coefficient for the pre-1932 period is not significantly negative. Data for installment debt are only available for years from 1919, which limits the utility of the cointegration-error correction model to test Olney’s avoiding default hypothesis. A variant of Olney’s own life-cycle model, which incorporates installment debt in different periods, and dummy variables to capture variation in the severity of debt default penalties, is adopted, and augmented as Eq. (2) to gauge the robustness of her model to the uncertainty hypothesis:
24 25 26 27 28
ln Ct = 0 + 1 ln Yt + 2 ln At + 3Vt + 4 ln ID2132, t1 + 5 ln ID3840, t1 + 1t
(2)
29 30 31 32 33 34 35 36 37 38 39 40
where ID21–32 is per capita real installment debt over the period from 1921 to 1932, and zero elsewhere, and ID38–40 is per capita real installment debt over the period from 1938 to 1940, and zero elsewhere. This formulation reflects that the severity of default penalties was ambiguous in the years 1933–1937.6 Note that the equation is estimated over the period from 1921 to 1940.7 This variant of the Olney model predicts that 4 < 0 and 5 > 0, and thus importantly that consumption falls with increased installment debt in the previous year, when the price of default is high. The data needed to estimate the augmented LCH-type consumption functions are described and discussed in the next section and the results are reported and discussed in Section IV.
4259 Ch02 4/12/02 9:43 am Page 61
Household Balance Sheet in U.S. Depression 1
61
III. THE DATA
2
Consumer Expenditures
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
The expenditure series used in the consumption function estimation are taken from Lebergott (1996). Lebergott’s data cover over one hundred items, and are consistent with and benchmarked to the U.S. Bureau of Economic Analysis (BEA) data in 1929. Most importantly, and in contrast to Shaw’s (1947) estimates used by Romer, these data cover both commodities and services. Components of Lebergott’s consumption data are allocated here between durables, semi-durables, perishables and services. To illustrate, elements of these data are presented in Table 1 for the period 1928–1933. While durables spending shows the sharpest falls, of around 50%, during the contraction, non-durable (perishables and services) spending accounts for 60% of the fall in total expenditures in the period 1929–1933. Although Mishkin and Romer focus on explaining the decline in durables spending, a satisfactory understanding of the consumption decline during the early years of the Great Depression also needs to account for shifts in non-durables expenditures. Services, according to Lebergott’s estimates, accounted for 51% of consumer spending in 1929. In addition to providing estimates of services’ spending,
21
Table 1.
22 23 24 25 26 27 28 29
Consumer Expenditures ($ billions, 1987 prices).
Year
Perishables
Services
Semi-durables
Durables
Total
1928 1929 1930 1931 1932 1933
164.4 168.3 164.8 164.4 154.1 151.1
265.7 280.2 265.2 254.1 233.6 231.1
59.9 61.7 56.1 53.6 46.9 42.7
42.4 44.6 34.8 30.0 22.5 22.6
532.3 554.8 520.8 501.9 456.9 447.6
30 31 32 33 34 35 36 37 38 39 40
Source: Lebergott (1996, Tables A2–A4). Numbers in parentheses refer to column numbers in Lebergott. Perishables. Total food, alcohol and tobacco (1), wood, gas and coal (34). Services: personal care (15), housing (18), domestic service (36), other household operation (37), medical care, except ophthalmic (38), personal business (44), auto repair (54), gas and oil for motor vehicles (55), auto installation and tolls (56), recreation, except books and maps, durable toys and wheel goods, and music, radio, and TV (59), education (67), electricity (31), gas (32), water (33), telephone and telex (35), religion and welfare (71), for traveling by U.S. residents (75), local transport (57), and intercity transport (58). Semi-durables. Clothing (7), semi-durable house furnish (28), clean and polish (29), stationary (30), and tires and accessories (53). Durables. Furniture and mattresses (24), kitchen appliances (25), china (26), other durable furniture (27), motor vehicles and wagons (52), and the recreation and medical components excluded from services (40), (60), (63), and (64).
61
4259 Ch02 4/12/02 9:43 am Page 62
62 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
DAVID GREASLEY AND JAKOB B. MADSEN
Lebergott adjusts Shaw’s commodity data to allow for changing distributive margins, and provides new estimates for food expenditures. His commodity spending estimates do not adjust Shaw’s output data for fluctuations in inventories. The utility of Lebergott’s data depends on the robustness of his estimates for 33 categories of services. Some of these data are direct estimates, including those for education and foreign travel, but other elements, including domestic and medical services, are based on employment and wages. In similar fashion to Shaw’s construction of commodity output data from occasional census returns, Lebergott interpolates certain categories of services’ spending, including legal services and intercity transportation. The advantage of his data lies principally in their breadth of coverage (they include services, the share of which in total spending grew from 41 to 51% between 1900–1929), and in the consistent benchmarking they provide to post-1929 BEA estimates. The disaggregated consumer spending data show a variety of relationships with disposable income during the early years of the Great Depression. The ratio of durables consumption to income, which had more than doubled in the period 1900–1929, fell around 30% between 1929–1933. Semi-durables consumption in relation to income also fell, by around 4% between 1929–1933, whereas the ratios of perishables and service consumption to income rose respectively 24% and 14%. Romer argues that in periods of heightened income uncertainty consumers switch their spending away from irreversible commitments to buy durables towards spending on perishables. Total consumer spending as a proportion of disposable income, however, rose around 11% over the years 1929–1933. Nevertheless, the decline in consumption may have led to a fall in income since contra-cyclical variations in the consumption-income ratio are an implication of the LCH. For consumers’ income uncertainty or deterioration in their balance sheets to have played a major role in the consumption decline at the onset of the Great Depression, the effects need to have extended beyond durable goods to perishables and services. Heightened uncertainty or balance sheet effects may have played important roles in reducing consumption, even for those categories where the ratio of spending to income rises over the period 1929–1933. Some categories of consumption will show counter-cyclical variation in relation to disposable income, when income decline is expected to be transitory. The central issue to gauge is whether or not the rise in perishable, service, and, indeed, total spending relative to income would have been substantially higher in the absence of heightened income uncertainty or deterioration in balance sheets. The model outlined in the previous section, which articulates both long run relationships between categories of consumption and income, and shorter-term variations in consumption, including those
4259 Ch02 4/12/02 9:43 am Page 63
Household Balance Sheet in U.S. Depression 1 2
63
in response to changes in households’ balance sheets and income uncertainty, permits investigation of the central issue.
3 4
Financial Assets and Liabilities.
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
Ando and Brown’s annual series for net wealth, which is based upon Goldsmith’s (1955) benchmarks, provides the starting point for measuring households’ financial assets and liabilities.8 Their aggregate data do not distinguish financial assets, non-financial assets, and liabilities. To disaggregate, a Mishkin-type measure is constructed where financial assets are calculated as a residual by subtracting tangible assets from and adding liabilities to net wealth.9 The measures used here for tangible assets and liabilities do not rest on interpolation, and the resultant financial assets data should better represent annual movements than Mishkin’s data, which are interpolated between 1929, 1933, and 1939 benchmarks. Our approach provides a measure of financial assets corresponding to items included by Goldsmith in the national balance sheet, including life insurance and pensions. In estimating financial assets, liabilities are defined as the sum of consumer mortgage and non-mortgage debt, and tangible assets as the sum of residential and consumer durable wealth. Data for non-mortgage consumer debt are available on an annual basis from 1900, and mortgage debt from 1916.10 The mortgage data are back-projected for earlier years using estimates of the housing stock’s value. Occasional pre-1916 data are available for mortgage debt, and these show a near constant relationship with the value of the housing stock. The ratio of mortgages to the value of residential and non-residential structures is in 1900, 15.98%, in 1912, 15.61%, and in 1922, 14.70% (Goldsmith, Lipsey & Mendelson, 1963).11 For tangible assets, annual data for market values of residential wealth are available, and a series for consumer durable wealth is constructed by applying the perpetual inventory method to the real consumption of durables using a 22% depreciation rate.12 Real consumer wealth is set equal to the average for durable expenditures in the first six years that data are available (1885–1890), multiplied by 1/0.22, which is the average life of a durable at the given depreciation rate. Table 2 compares the growth in our estimates of financial assets, and liabilities, deflated by consumer prices, with Mishkin’s estimates. The disparities highlight the possible weaknesses of using interpolation during a period of macroeconomic instability. The discrepancies are most pronounced for liabilities. Mishkin’s interpolated figures show a near 20% increase in 1930, while our estimate shows an increase of 0.5%.13 Mishkin’s data showing a sharp rise in real liabilities in 1930 appear implausible. Olney reports nominal installment debt fell 12.4% in 63
4259 Ch02 4/12/02 9:43 am Page 64
64 1
DAVID GREASLEY AND JAKOB B. MADSEN
Table 2.
2
Changes in Households’ Real Financial Assets and Liabilities (percent).
3 Financial Assets (Mishkin)
4 5 6 7 8 9 10 11 12
1930 1931 1932 1933
Financial Assets
3.9
1.7
5.6
8.1
7.8
8.7
2.2
6.1
Liabilities (Mishkin) 19.6 1.8 1.9 6.3
Liabilities 0.5 2.9 4.1 6.7
Notes: Mishkin (1978, p. 920) is the source of his data. Consumer prices are used as the deflator. Our measure of financial assets reported here incorporates Lebergott’s (1996) consumer durables data.
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
1930, and consumer prices fell only around 3% in that year. Nominal mortgage debt rose around 2.5% in 1930, which explains why our estimate of real debt rises slightly in 1930. Part of the discrepancy may arise from Mishkin’s use of beginning year values for debt. The empirical models estimated here do include previous year debt, and thus allow for the possibility that sharp rises in debt in 1929 diminished consumption in 1930. Mishkin’s interpolated data for financial assets, although less at variance with our estimates, show values which fall continuously between 1929 and 1932. Our measure shows a rise in the real value of financial assets in 1930. Part of the difference arises from direct measures for tangible assets used in our constructions. Lebergott’s estimates of consumer durables estimates were not available to Mishkin. The direct estimates of the market value of residential wealth show a fall in nominal terms of around 25% between 1929–1933. These data are constructed by adding increments each year of net residential capital formation to an initial year value. This approach should indicate more reliably annual movements than interpolation between benchmarks, although the differences are not great.
31 32
Uncertainty
33 34 35 36 37 38 39 40
Different approaches have been used for measuring consumers’ income uncertainty. Most indicators are based on the volatility of prices (including interest rates and share prices), output, or exchange rates. Wilson, Sylla and Jones (1990) discuss alternative measures of share price volatility. That used by Romer, and deployed here, as a proxy for consumers’ income uncertainty is the standard deviation of monthly share prices, averaged over 12 months to September, and illustrated in Fig. 1.
4259 Ch02 4/12/02 9:43 am Page 65
Household Balance Sheet in U.S. Depression 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
65
Most debate surrounding the use of share price volatility to measure preWorld War Two income uncertainty concerns the low numbers, around 2% of Americans in 1929, holding shares. Romer, though, argues that the stock market can be considered an imperfect predictor of the real economy, citing the views of forecasters in the financial press of 1929. Flacco and Parker’s (1992) econometric estimates of income variance for the 1920s, which they use as a measure of income uncertainty, support her position.14 They conclude that the unprecedented negative shocks to the stock market dominate their statistical measure of income variance in late 1929. Further, they show that the decline of one measure of consumption, department store sales, at the onset of the Great Depression, can be explained substantially by their income variance estimate of uncertainty. Flacco and Parker replicated their analysis of consumption using share price volatility as an alternative to their statistical estimate of income variance, and obtained similar results. They conclude that the stock market volatility measure seems to be a good proxy for income uncertainty during the 1920s. For later years there is supporting statistical evidence that consumers generally are influenced by stock market events. The findings of Otoo (1999), and Madsen and McAleer (2000) for the post-1945 period show that consumer confidence is strongly related to share prices. Most importantly, Otoo found that share price movements influenced consumer confidence among owners and non-owners of shares equally. For the 1920s there are good reasons to believe that the uncertainty spawned by share price volatility affected not only business forecasters and holders of stock but consumers more generally. The public’s attention to the stock market grew strongly during the bull market in the 1920s. This was an inevitable consequence of increased coverage by the press and by the periodical literature. Shiller (2000), for example, shows that the percentage of articles in the Reader’s Guide to the Periodical Literature dealing with the stock market quadrupled during the 1920s. Galbraith (1955) goes further, arguing that by the summer of 1929, only the most aggressive of eccentrics maintained a detachment from the stock market. Consumers, generally, were probably aware that events in the stock market during 1929 were unprecedented. For the years to 1914, Wilson, Sylla and Jones (1990) argue that banking panics and the stock market crashes that accompanied them occurred in the USA every ten to twenty years, and they maintain that these events, though unsettling, were not especially surprising. In contrast, they contend that the stock market crash of 1929, in the era of the Federal Reserve, was different for three reasons. First, the extent of the share price decline in October and November 1929 was unprecedented relative to any other two month decline since 1834. Second the stock market 65
4259 Ch02 4/12/02 9:43 am Page 66
66
DAVID GREASLEY AND JAKOB B. MADSEN
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Fig. 1.
16 17
Real Share Price Volatility.
Source: Romer (1990).
18 19 20 21 22 23 24 25 26 27 28 29
crash was unrelated to a banking panic. And third, stock price volatility persisted, and rose to new heights over the following three years. The stock market crash captured the imagination of the public generally, and its unprecedented characteristics meant its consequences were uncertain, for households and businesses. Figure 1 illustrates the measure of share price volatility used to indicate income uncertainty in the results section. This measure of income uncertainty shows a sharp rise to 0.009 in the year to September 1930, compared to its average pre-1929 value of 0.001, illustrated as the horizontal line in Fig. 1. The stock volatility index falls to 0.006 in the year to September 1931, before peaking at 0.026 in the year to September 1932.
30 31
IV. EMPIRICAL MODELS
32 33
Estimation Results.
34 35 36 37 38 39 40
The section reports results from the life-cycle type consumption functions formulated within a modern time series framework. All variables were first tested for order of integration using augmented Dickey-Fuller tests and the results are reported in Table 3. The table shows that all variables are integrated of order 1 at the 10% level, except share price volatility and the real interest rate, which are stationary variables, and disposable income. The tests for disposable income give
4259 Ch02 4/12/02 9:43 am Page 67
Household Balance Sheet in U.S. Depression 1
Table 3.
67
Augmented Dickey-Fuller Tests for Unit Root.
2 logs, no trend
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Disposable income Financial assets Liabilities Relative price of durables Net wealth Share price variability Installment debt Deposits of suspended banks Real interest rate Consumption, total Consumption, durables Consumption, perishables Consumption, semi-durables Consumption, services
logs, trend
log first difference, no trend
3.22
3.24
3.49
1.63
2.03
6.31
2.23
2.30
3.86
1.14
1.98
2.63
1.57
1.80
6.38
3.54
3.81
+ 0.04 2.29 4.49 2.53 1.21 1.71 2.56 1.39
1.38
2.86
2.13
6.51
4.61
–
–
3.29
2.72
1.33
2.44
1.73
3.66
2.53
4.25
2.51
2.96
Notes: The critical values at the 10% level are: 2.57 in logs with no trend, 3.13 in logs and time-trend included, and 2.57 in log first differences and with no time-trend included. The number of lags in the augmented Dickey-Fuller tests are based on the Akaike Information Criteria. Estimation period: 1902–1939. Sources: See Section III, and for deposits of suspended banks, Historical Statistics, TablesX741–67. The real interest rate is the nominal short-term interest rate minus the expected inflation rate based on prices of durables (see footnote 30 for data sources). The expected inflation is generated as a forecast of an ARMA (2,2) model. The short interest rate is Prime bankers’ acceptances, 90 days, Historical Statistics, Table X449.
24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
ambiguous results, and in the estimations below it is treated as an integrated variable. The implication of the stationarity tests is that the models can be formulated in error-correction form, such as Eq. (1). Equation (1) is expanded with one-period lags of all first-difference variables lagged one period and estimated using the general-to-specific estimation strategy where insignificant variables are sequentially deleted. A 10% significance level is used. The unrestricted and restricted estimation results for durables spending are shown in Table 4. The results for other categories, and for total spending shown in Table 5 simply are for the restricted model following the sequential deletion of the statistically insignificant variables. The diagnostic tests do not give evidence against the model specifications. The statistical insignificance of the error correction term implies, for each category of spending, that shifts in consumption were not influenced by the deviation of the ratio of consumption to income from its long-run equilibrium. 67
4259 Ch02 4/12/02 9:43 am Page 68
68
DAVID GREASLEY AND JAKOB B. MADSEN
Table 4.
1 2
Unrestricted and Restricted Parameter Estimates of Durable Consumption.
3 4
ln (Y)t
–0.94 (0.48)
5
ln (Y)t1
–0.46 (0.54)
6
ln (FAS)t
0.28 (1.27)
7
ln (FAS)t1
8
ln (LIA)t
9
ln (LIA)t1
10 11 12 13 14 15 16 17 18
0.51 (2.69) 0.52 (1.20)
Vt
–9.70 (2.02)
Vt1
–8.23 (1.14)
ln (P du /Pt) t
–1.08 (2.36)
/Pt1) ln (P du t1
–0.07 (0.15)
RRt
0.79 (2.01)
RRt1
0.15 (0.40)
ln (SUSt)
0.01 (0.24)
ln (SUSt1)
0.45 (3.18)
–0.29 (1.13) –8.30 (2.39) –1.90 (5.75)
–0.04 (1.58)
ln (Ct1 /Yt1)
0.07 (0.24)
ln (Ct1 /At1)
–0.07 (0.23) –0.12 (0.23)
22
Constant ¯2 R
23
DW
24
B-P
25
RESET
0.58
0.34
26
ARCH
0.23
0.12
19 20 21
27 28 29 30 31 32 33 34 35 36 37 38 39 40
0.65 1.82 16.4
–0.01 (0.40) 0.62 2.20 3.79
Notes: RR is the real interest rate and SUS is deposits in suspended banks. t-statistics are in ¯ 2 = adjusted multiple correlation coefficient. DW = Durbin-Watson statistics for firstparentheses. R order serial correlation. B-P = Breusch-Pagan test for heteroscedasticity, where the squared residual is regressed on original regressors, and is distributed as 2(k) under the null hypothesis of no heteroscedasticity, where k is the number of regressors excluding the constant term. RESET = RESET test for functional form, adding the predicted value squared to the regression, and is distributed as F(1, nk + 2) under the null hypothesis of correct functional form, where k is the number of regressors in the original regression excluding the constant term and n is the number of observations (39). One period lags of the following variables measured in log first differences are used as instruments for ln Yt: per capita real disposable income, manufacturing wages, consumer prices, and M2 deflated by consumer prices. Estimation period is 1902 to 1939. Consumer prices and M2 are from Historical Statistics, Tables E135 (consumer prices) and X 415 (M2). Disposable income is from Lebergott (1996). Manufacturing wages are from Williamson (1995) and for after 1919 from Historical Statistics, Table D802.
4259 Ch02 4/12/02 9:43 am Page 69
Household Balance Sheet in U.S. Depression 1
Table 5.
2 3 5 6 7 8 9 10 11 12 13
Parameter Estimates of Semi- and Non-Durables Consumption Functions. Semi-durables
4 ln Ct1 ln Yt Vt Vt1 ln FASt Constant ¯2 R DW B-P RESET ARCH
69
Services
Perishables
Total
0.43 (3.14) 5.29
(2.64) 6.77 (3.23) 0.26 (2.02) 0.00 (0.49) 0.32 2.04 6.99 0.50 0.73
3.41
(3.16)
0.11 (1.77) 0.00 (0.94) 0.29 1.88 3.11 0.91 0.11
1.35 (4.49) (1.92)
2.64
0.11 (2.08) (1.86) 0.45 1.88 1.86 2.36 0.45
0.01
0.69 (2.67) (3.49) 2.93 (2.42) 4.19
0.00 (0.26) 0.46 1.84 2.10 2.18 0.03
14 15
Notes: as for Table 4.
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
This suggests that a long-run relationship between consumption, income and wealth is not present in the sample period considered here.15 Since the variables are not cointegrated, lnY is not exogenous, and is instrumented, as reported in the notes to Table 4. The estimated coefficient of the uncertainty variable is significant in all the estimates and therefore provides robust support for the uncertainty hypothesis.16 After controlling for the possible effects of shifts in consumption towards equilibrium levels determined by income and wealth, as well as for shorterterm fluctuations in income, consumers’ income uncertainty is shown to significantly affect all categories of their spending. The influence of uncertainty appears strongest in the case of durables consumption, with the size of the estimated coefficient for the effects of the change in current year uncertainty over twice that for total consumption. The findings provide only partial support for the liquidity hypothesis. Changes in households’ holdings of financial assets are shown to statistically significantly affect changes in all categories of consumption other than services, but as is shown below, the quantitative importance of this influence appears modest. Indeed, Mishkin’s own model applies strictly to durables spending, and lays most emphasis on the impact of debt on the downturn on consumer durables spending at the onset of the Great Depression. Our empirical models do not support the hypothesis that changes in households’ liabilities influenced durables consumption. Further, for no categories of spending do the results show households’ debts to significantly influence consumption. The simple liquidity 69
4259 Ch02 4/12/02 9:43 am Page 70
70 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
DAVID GREASLEY AND JAKOB B. MADSEN
hypothesis does not appear robust to the inclusion of income uncertainty within a unified consumption model. Olney, though, refines the liquidity model to take account of variations in the severity of default penalties on installment debt. The results from estimating Eq. (2) are shown in Table 6.17 The estimates show that adding a measure for income uncertainty to an Olney-type model undermines further the liquidity hypothesis. The debt default variant of the liquidity model postulates negative coefficients for lagged installment debt prior to 1932 when penalties were severe, and positive ones from 1938, when they were not. The results do not show statistically significant negative coefficients for pre-1932 installment debt, or significant negative coefficients for 1938–1940, for any category of spending. Indeed, for total spending, higher previous year installment debt before 1932 appears associated with higher spending in the current year. In contrast, the uncertainty hypothesis receives further support from the results in Table 6, which incorporate the effects of changes in income and wealth on consumption, and show for all categories of consumption that heightened income uncertainty led to diminished spending. Olney’s interpretation that a diminished supply of credit was not central to the collapse of consumer durables spending in 1930–1932 receives support from the results in Table 4. The indicator, the deposits of suspended banks, does
21 22
Table 6.
Consumption, Income Uncertainty and Installment Debt.
23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
ln Yt ln (P du /Pt) t ln At Vt Vt1 ID2132, t1 ID3840, t1 Constant R¯ 2 DW B-P RESET ARCH
Durables
Semi-durables
4.55 (3.86) –0.44 (2.98)
0.66 (2.44)
–14.3 (5.41) 0.22 (0.95) –5.28 (2.81) –0.07 (2.42) 0.54 1.60 39.70 19.30 0.01
0.61 –4.47 –7.30 –0.05 0.06 0.01 0.53 1.73 30.7 20.87 5.10
(3.00) (2.15) (3.89) (0.36) (0.49) (0.65)
Services
Perishables
0.92 (3.36)
1.21 (4.59)
–3.66 (3.83) 0.20 (2.55) 0.00 (0.03) 0.00 0.99) 0.39 1.82 39.85 17.02 0.03
0.28 –2.67 –2.45 –0.03 0.03 –0.01 0.63 1.58 37.13 24.67 1.69
(3.19) (3.30) (2.51) (0.58) (0.41) (0.63)
Total
–4.62 (4.42) 0.18 (1.93) –0.01 (0.20) –0.01 (1.36) 0.49 1.51 39.86 16.38 0.03
Notes: See notes to Table 4. Estimation period 1921 to 1940. The t-values are based on White’s heteroscedasticity consistent covariance matrix.
4259 Ch02 4/12/02 9:43 am Page 71
Household Balance Sheet in U.S. Depression 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
71
not exhibit a statistically significant relationship with durables spending. Furthermore, the banks did not supply much installment credit before the great contraction. Their share of the total lending was only around five percent in 1929, and thus this statistical finding appears unsurprising. Moreover, the severity of default penalties (missing a single payment could lead to uncompensated repossession of the durable good by the finance company) diminished the relevance of adverse selection. Olney (1991) notes that finance companies, in stark contrast to the banks, were largely unaffected by bankruptcies.18 That neither debt nor credit restrictions contributed significantly to the collapse of consumer spending at the onset of the Great Depression strengthen greatly the likely central role played by uncertainty surrounding incomes. Finally, the results for durable consumption in Table 4 indicate that the real interest rate does not have the expected negative effect, where the real interest rate is measured as the nominal rate on a short-term government bond minus an ARMA(2,2) forecast of the rate of inflation in prices of durables. The estimated coefficients of the real interest rate were positive. Sensitivity analysis using other measures of the real interest rate did not change this result.19 The estimated coefficients were statistically insignificant and positive.
19 20 21
Simulation Results and Implications
22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
The respective effects of the uncertainty and balance sheet variables on consumption at the onset of the Great Depression are gauged by simulating their estimated effects, utilizing the results in Tables 4 and 5, and focusing especially on 1930 and 1932, the years of largest spending declines.20 For total consumption in 1930, the estimated effect of heightened income uncertainty as measured by share price volatility reduced per capita spending by 3.23%, compared to the actual fall of 7.47%. Since real financial asset depletion did not occur until 1931–1932 (see Table 2), the deterioration on households’ balance sheets had its largest estimated effect in 1932, reducing total consumption per capita in that year by 0.90%. In contrast, income uncertainty is estimated to have reduced total consumption per capita in 1932 by 7.47%, compared to the actual fall of 10.03% in that year. Neither the depletion of households’ financial asset nor income uncertainty had substantial effects in 1931. Overall spending per capita fell 4.47% in that year, with the effects of reduced income estimated to have diminished spending by 4.09%. Income uncertainty was the dominant reason for the falls in total consumption in 1930 and 1932. A comparison with the experience of 1921 helps inform what distinguished the Great Depression. In 1921 when total consumption per 71
4259 Ch02 4/12/02 9:43 am Page 72
72 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
DAVID GREASLEY AND JAKOB B. MADSEN
capita fell by 6.59%, income uncertainty was unimportant. The simulation results show that a depletion of financial assets and the fall in income respectively reduced spending per capita by 1.95 and 3.70% in 1921, while income uncertainty is estimated to have reduced consumption by 0.21%. On these results, what set the onset of Great Depression apart was the rise in consumers’ income uncertainty, which, according to the estimates, explain 43% of the fall in overall consumption per capita in 1930. While uncertainty, rather than deterioration in households’ balance sheets, appears central to understanding the total consumption collapse at the onset of the Great Depression, the continuing fall in consumer spending through 1931 resulted largely from the fall in income. However, a further heightening of income uncertainty in 1932 lies at the heart of the Great Depression’s most substantial falls in consumption, which occurred in that year. The results for both categories of non-durables mirror those for total spending. Thus, income uncertainty is estimated to have reduced services spending per capita by 2.39 and 6.74% in 1930 and 1932 respectively, while the comparable figures for perishables are 1.85 and 5.22%. The depletion of financial assets is estimated to have reduced spending on perishables modestly, by less than 1% in 1931 and 1932, and to have had no effect on services spending. For durables consumption, the position is complicated by the estimated effects of higher real prices for this category of consumption, especially during 1931–1933. In 1930 and 1932 income uncertainty is estimated to have reduced durables spending per capita by 5.84 and 16.45% respectively. Higher real prices for durables are estimated to have reduced their consumption per capita by 1.09% in 1930, and the fall in income reduced durables consumption in the same year by 6.39%. Actual durable spending per capita fell by 15.56% in 1931, and the fall in households’ financial assets in the previous year and rising durables prices are estimated to have respectively reduced spending in 1931 by 3.7 and 9.14%. In contrast, the further heightening of income uncertainty in 1932 is estimated to have reduced durables spending per capita by 16.45% in that year, compared to the actual fall of 29.48%. The results for semi-durables are somewhat distinct from those for other categories of spending. Semi-durables consumption per capita fell steadily throughout the years 1930–1933. For example, the fall of 9.80% in 1933 was almost as severe as the 10.82% fall in this category of spending in 1930. The estimates show that heightened income uncertainty explain 40% of the fall in semi-durable spending per capita in 1930, and between 60–66% of the falls between 1931–1933. The depletion of households’ financial assets contributed to the spending falls for this category of consumption in 1931–1932, but not in 1930 or 1933.
4259 Ch02 4/12/02 9:43 am Page 73
Household Balance Sheet in U.S. Depression 1 2 3 4 5 6 7 8 9 10 11 12
73
Overall, the results in Tables 4 and 5 point clearly to uncertainty rather than illiquidity depressing consumer spending at the onset of the contraction. For all categories of consumers’ expenditures, income uncertainty as measured by share price volatility is shown to reduce spending. Since share price volatility rose most in 1930 and 1932, this measure of uncertainty had greatest effect in these years, although substantial lagged effects, and thus impact in 1931 and 1933, were found in the case of semi-durables. Depletion of households’ financial assets did contribute to the spending falls of 1931–1932, but except in the case of semi-durables, the effect was modest. For durables, the rise in the relative price of this category of spending had a more deleterious effect on consumption, especially between 1931–1933, than the deterioration of households’ balance sheets.
13 14
V. CONCLUSION
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
Previous attempts to explain the falls in consumption during the early years of the Great Depression focus separately on the household balance-sheet, especially on the role of debt, the supply of credit, and on uncertainty, and pay most attention to 1930 and to durable spending. This article considers the various hypotheses within a unified consumption model, and gives greater attention to non-durable spending, and to the events of 1931–1933. Our key findings show that income uncertainty precipitated and prolonged the consumption declines at the onset of the Great Depression. Reduced holdings by households of financial assets contributed to the spending declines of 1931–1932, but the magnitude of the adverse balance sheet effect was modest for total spending, and does not contribute to explaining the spending falls of 1930. Nor does excessive indebtedness, or reduced credit supplies, from the evidence of our estimates, explain the consumption downturn at the onset of the Great Depression. Mishkin and Olney have highlighted the role of debt, but neither version of the hypothesis appears robust. The rise in the real value of households’ liabilities, broadly defined, was lower in 1930 than Mishkin contends, and did not significantly influence durables consumption. Nor do the results here support Olney’s view that the unprecedented rise in installment debt, and the severe default penalties prior to 1932, explain most of the consumption collapse in 1930. No difference was found here between the estimated impact of debt on consumption in years to 1932, when default penalties were severe, and after 1938, when they were not, after incorporating the effects of uncertainty. Heightened income uncertainty was chiefly responsible for the collapse in all categories of consumer spending in 1930 and 1932. Uncertainty was less important in 1931. The pace of the consumption decline ameliorated in 1931, 73
4259 Ch02 4/12/02 9:43 am Page 74
74 1 2 3 4 5 6 7 8 9 10 11 12
DAVID GREASLEY AND JAKOB B. MADSEN
and was driven principally by income falls and increases in the real price of durables. The prolongation of the contraction through 1932, though, was associated with a further rise in consumers’ income uncertainty. Our interpretation postulates that consumers increased their precautionary savings by reducing spending at times of uncertainty surrounding their future income. Additionally, for durables, higher prices and the irreversibility of purchases once made, reinforced the effects of a rise in precautionary savings. The force of income uncertainty was in 1930 and 1932 sufficient to account for most of the collapse in total spending, irrespective of the levels of households’ debt. Uncertainty was reduced greatly in 1933. By 1935 consumers’ certainty had recovered to the extent that non-mortgage debt had risen to a greater proportion of average incomes than in 1929, and households’ spending grew strongly.
13 14
NOTES
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
1. Although Mishkin (1990) does not focus explicitly on consumption its content has relevance for the debates surrounding durables spending. 2. In the case of the life cycle model the unexplained falls in total consumption are 9% and 7%, respectively, in 1932 and 1930, while the comparable figures for non durables are 6% and 4%, respectively. 3. Bernanke (1983a, pp. 269–270) utilizes monthly data and makes adjustments for the bank holiday of March 1933. Our annual data do not include a similar adjustment since the deposits of banks which were re-opened by June 1933 are not included in the suspensions data. 4. If the amount of consumption the consumers are willing to give up to avoid a given amount of uncertainty, is an increasing function of wealth, then consumption is a declining function of income uncertainty. That is, the marginal utility of consumption is a convex function of consumption, which implies that the marginal utility of consumption is a declining function of consumption but at a declining rate. The convexity of the marginal utility implies that marginal utility of consumption is higher for the average consumption than the sum of the marginal utilities for various values of consumption. More formally income uncertainty lowers consumption if the third derivative of a von Neumann-Morgenstein utility function is positive. 5. This is shown by Black and Scholes (1973). 6. Courts in several states deemed default penalties unduly harsh during these years, but the legal landscape nationwide was not changed until 1938. See Olney (1999, p. 327). 7. Error correction terms analogous to those in Eq. (1) were incorporated in initial estimates but were found statistically insignificant. In light of the modest sample size, and that Olney’s own model does not include error correction terms, they are not shown in the specification of Eq. (2). 8. Ando and Brown’s data are reported in Modigliani (1966, pp. 215–216). 9. Details of Mishkin’s procedure are in Mishkin (1978, appendix). 10. Non-mortgage consumer debt is from Olney (1991) for years to 1919, and thereafter from Historical statistics, Table X551. For years from 1916 mortgage debt is from Historical statistics, Tables X393–409.
4259 Ch02 4/12/02 9:43 am Page 75
Household Balance Sheet in U.S. Depression 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
75
11. The non-residential structures included here are those defined as part of household wealth, and include, for example, garages. 12. For residential wealth see, Grebler, Blank and Winnick (1956) Consumer durable spending is from Lebergott (1996) for 1900–1941, and Kendrick (1961) for 1885–1899. The consumer durable deflator is from Balke and Gordon (1986) for 1919–1941, for the last quarter of each year, and Historical statistics, Tables P 318–374 for 1889–1918. 13. Mishkin has only four data points for the interwar years, 1922, 1929, 1933, and 1939. He uses a combination of linear and constant growth approaches to interpolation. These methods are unlikely to represent accurately annual movements in households’ assets and liabilities. 14. In the absence of monthly data for disposable income they use industrial production as a proxy. The correspondence between output variance and share price volatility adds credence to the view that stock market volatility can be considered an imperfect predictor of the real economy. 15. The absence of a long-run relationship is likely to reflect the small sample and macroeconomic instability. To test for the possibility that the cointegration tests are sensitive to the choice of estimator, the tests of Johansen were undertaken using three lags of all variables. The null hypothesis of no cointegration vectors could not be rejected at the 10% level. 16. Since the consumption of housing services, which formed around 20% of services spending, cannot easily be varied, the regression for services was also run excluding housing services, but the results of Table 4 were not materially affected. 17. The diagnostic tests indicate heteroscedasticity and functional form problems for all consumer categories and in some cases also the possibility of first order serial correlation, thus suggesting that the Olney-type model is mis-specified. Although White’s heteroscedasticity consistent covariance matrix is used, the parameter estimates may be biased. 18. She reports evidence that only five banks suffered defaults on loans to finance companies in the period 1930–1935. 19. The estimates were based on different measures of the interest rate and different measures of inflation expectations. The estimates included both the short interest rate, as in the estimates above, and the interest rate on a long-term government bond. Inflation expectations were measured as the actual inflation rate, a one-period lag of the inflation rate and an AR(1) forecast of the inflation rate which includes a lagged dependent variable, lagged growth in M2 and the lagged short-term interest rate. Inflation rates were based on prices of durable goods. 20. The simulations are dynamic in the case of services to allow previous year uncertainty via its effect on previous year consumption to influence current year consumption. The results in Table 5 show shifts in lagged consumption are statistically significant only in the case of services spending.
35 36
ACKNOWLEDGMENTS
37 38 39 40
Comments and suggestions by a referee, and by Les Oxley, are gratefully acknowledged. The paper was written while the second author was employed at the University of Western Australia. 75
4259 Ch02 4/12/02 9:43 am Page 76
76 1
DAVID GREASLEY AND JAKOB B. MADSEN
REFERENCES
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
Ando, A., & Modigliani, F. (1962). The Life Cycle Hypothesis of Savings: Aggregate Implications and Tests. American Economic Review, 53, 53–84. Balke, N. S., & Gordon, R. J. (1980). Appendix B. Historical Data. In: R. J. Gordon (Ed.), The American Business Cycle: Continuity and Change (pp. 781–850). Chicago: University of Chicago Press. Bernanke, B. S. (1983a). Irreversibility, Uncertainty and Cyclical Investment. Quarterly Journal of Economics, 98, 85–106. Bernanke, B. S. (1983b). Non-Monetary Effects of Financial Crises and the Propagation of the Great Depression. American Economic Review, 73, 257–276. Bernanke, B. S. (1985). Adjustment Costs, Durables, and Aggregate Consumption. Journal of Monetary Economics, 15, 41–68. Black, F., & Scholes, M. (1973). The Price of Options and Corporate Liabilities. Journal of Political Economy, 81, 637–654. Brown, C. (1997). Consumer Credit and the Propensity to Consume: Evidence From 1930. Journal of Post Keynesian Economics, 19, 617–638. Cowles, A., & Associates (1939). Common-stock indexes. Bloomington: Principa Press. Eichengreen, B. (1992). The Origins and Nature of the Great Slump Revisited. Economic History Review (2nd series XLV), 213–239. Engle, R. F. (1982). Autoregressive Regressive Conditional Heteroscedasticity With Estimates of the Variances of United Kingdom Inflations. Econometrica, 50, 987–1007. Ferderer, P., & Zalewski, D. (1994). Uncertainty As a Propagating Factor in the Great Depression. Journal of Economic History, 54, 825–849. Flacco, P., & Parker, R. (1992). Income Uncertainty and the Onset of the Great Depression. Economic Inquiry, 30, 154–171. Friedman, M., & Schwartz, A. J. (1963). A Monetary History of the United States, 1867–1960. Princeton: Princeton University Press. Galbraith, J. K. (1955). The Great Crash. London: Penguin. Goldsmith, R. W. (1955). A Study of Savings in the United States. Princeton: Princeton University Press. Goldsmith, R. W., Lipsey, R. E., & Mendelson, M. (1963). Studies in National Balance Sheet of the United States. Princeton: Princeton University Press. Grebler, L., Blank, D. M., & Winnick, L. (1956). Capital Formation in Residential Real Estate, Trends and Prospects. Princeton: Princeton University Press. Greasley, D., Madsen, J. B., & Oxley, L. (2001). Income Uncertainty and Consumer Spending During the Great Depression. Explorations in Economic History, 38, 225–251. Fearon, P. (1987). War Prosperity and Depression: 1917–1945. Oxford: Philip Allan. Hahm, J.-H., & Steigerwald, D. (1999). Consumption Adjustment Under Time-Varying Income Uncertainty. Review of Economic and Statistics, 81, 32–40. Hall, R. E. (1978). Stochastic Implications of the Life Cycle-Permanent Income Hypothesis: Theory and Evidence. Journal of Political Economy, 86, 971–987. Hamburger, M. (1967). Interest Rates and the Demand for Consumer Durable Goods. American Economic Review, 57, 1131–1153. International Monetary Fund (various years). International Financial Statistics. Washington. Kendrick, J. W. (1961). Productivity Trends in the United States. Princeton: Princeton University Press.
4259 Ch02 4/12/02 9:43 am Page 77
Household Balance Sheet in U.S. Depression 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
77
League of Nations (various years). Monthly Bulletin of Statistics. Geneva. Lebergott, S. (1996). Consumer Expenditures: New Measures and Old Motives. Princeton: Princeton University Press. Leisner, T. (1989). One Hundred Years of Economic Statistics. Oxford: Economist Publications Ltd. Madsen, J. B., & McAleer, M. (2000). Direct Tests of the Permanent Income Hypothesis Under Uncertainty, Inflationary Expectations and Liquidity Constraints. Journal of Macroeconomics, 22, 229–252. Mankiw, N. G. (1985). Consumer Durables and the Real Interest Rate. Review of Economic and Statistics, 62, 353–362. Mayer, T. (1978). Consumption in the Great Depression. Journal of Political Economy, 86, 139–145. Mishkin, F. S. (1976). Illiquidity, Consumer Durables Expenditure, and Monetary Policy. American Economic Review, 66, 642–654. Mishkin, F. S. (1978). The Household Balance Sheet and the Great Depression. Journal of Economic History, 38, 918–937. Mishkin, F. S. (1990). Asymmetric Information and Financial Crises: A Historical Perspective. National Bureau of Economic Research Working Paper No. 3400, Cambridge, MA. Modigliani, F. (1966). The Life Cycle Hypothesis of Saving, the Demand for Wealth and the Supply of Capital. Social Research, 33, 160–217. Olney, M. L. (1991). Buy Now, Pay Later. Chapel Hill: University of North Carolina Press. Olney, M. L. (1999). Avoiding Default: The Role of Credit in the Consumption Collapse of 1930. Quarterly Journal of Economics, 114, 319–335. Otoo, M. W. (1999). Consumer Confidence and the Stock Market. Working Paper, Board of the Governors of the Federal Reserve System. Romer, C. (1990). The Great Crash and the Onset of the Great Depression. Quarterly Journal of Economics, 105, 597–624. Shaw, W. (1947). Value of Commodity Product Since 1869. New York: NBER. Shiller, R. J. (2000). Irrational Exuberance. Princeton: Princeton University Press. Temin, P. (1976). Did Monetary Forces Course the Great Depression? New York: Norton. United States Bureau of Census (1975). Historical Statistics of the United States: From Colonial Times to the Present. Washington: Government Printing Office. Warren, G. F., & Pearson, F. A. (1932). Wholesale Prices in the United States for 135 Years, 1797–1932. New York: Cornell University Agricultural Experiment Station Memoir 142. Wilson, J. W., Sylla, R. E., & Jones, C. P. (1990). Financial Market Panics and Volatility in the Long Run, 1830–1988. In: E. White (Ed.), Crashes and Panics: The Lessons of History. Homewood: Solomon Brothers Center for the Study of Financial Institutions. Williamson, J. G. (1995). The Evolution of Global Labor Markets Since 1830: Background Evidence and Hypothesis. Explorations in Economic History, 32, 141–196.
33 34 35 36 37 38 39 40
77
4259 Ch03 4/12/02 9:44 am Page 79
1 2 3 4 5 6 7 8 9 10
COPING WITH FINANCIAL CATASTROPHE: THE SAN FRANCISCO CLEARINGHOUSE DURING THE EARTHQUAKE OF 1906
11 12 13 14
Ronnie J. Phillips
15 16 17 18
ABSTRACT
19 20 21 22 23 24 25 26 27 28
The assumption today is that the Federal Reserve stands behind the financial system in case of a catastrophic shock. There has been little research on how the payments system functioned during economic catastrophes prior to the establishment of the Federal Reserve System. This paper examines the 1906 San Francisco earthquake when a private sector response was required after disaster occurred. The research question addressed is how well the private sector responded when there was a large external shock to the payments system such as an earthquake. The San Francisco Clearinghouse is examined as a case study.
29 30
I. INTRODUCTION
31 32 33 34 35
Today the bank regulatory structure that includes deposit insurance through the FDIC and a lender of last resort function through the Federal Reserve provides for the ultimate safety and stability of the financial system. In the event of a natural or financial “shock” to the system, it is presumed that these agencies
36 37 38 39 40
Research in Economic History, Volume 21, pages 79–104. Copyright © 2003 by Elsevier Science Ltd. All rights of reproduction in any form reserved. ISBN: 0-7623-0993-8
79
4259 Ch03 4/12/02 9:44 am Page 80
80 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
RONNIE J. PHILLIPS
will enable the financial system to continue to function. Thus, while many people were concerned about the Y2K problem, they were reassured that the Federal Reserve was printing up a few extra billion in Federal Reserve Notes. But prior to the establishment of this regulatory structure, what happened to the financial system when there were shocks such as earthquakes and panics? How did the private sector respond? Did the system sink into chaos, or were normal operations resumed quickly? Though there have been studies of panics under the national banking system, mostly for Chicago and New York, the experience of the other clearinghouses in crisis times has not been recently examined (Dwyer & Gilbert, 1989; Tallman & Moen, 1995; Roberds, 1995). The purpose of this paper is to supplement knowledge of the actions of a clearinghouse during shocks by examining the experience of the San Francisco Clearinghouse during the earthquake of 1906. Coordinated actions by clearinghouses in the U.S. in the nineteenth century represented a private sector solution to the problems inherent in fractional reserve banking. They were attempts to mitigate the effects of weaknesses of the National Banking System and establish a safe and stable banking system. Clearinghouses are an example of Alfred Chandler’s “Visible Hand” whereby the market mechanism is supplanted by management by American businessmen, in this case bankers, in coordinating the activities of the economy (Chandler, 1977, p. 1). The Federal Reserve System, established in 1913, was itself based on clearinghouse experience (Timberlake, 1993). The next two sections of the paper cover the early history of California banking and the role of clearinghouses in the nineteenth and early twentieth centuries. This is followed by a detailed review of the actions of the San Francisco clearinghouse during the earthquake of 1906. The paper concludes with the lessons of history for the operation of the payments system.
28 29 30
II. CALIFORNIA BANKING AND THE SAN FRANCISCO CLEARINGHOUSE
31 32
Early Banking in California
33 34 35 36 37 38 39 40
The California constitution prohibited the state from “granting any charter for banking purposes” (Article IV, Section 34). The banks that operated were private institutions that survived on reputation (Chandler, 1986, p. 3). During the early Gold Rush days, the bankers provided two fundamental services: paying miners in gold coin for the gold dust, and selling the drafts drawn on Eastern banks as a means of sending money through the mail (Chandler, 1986, p. 3; Wright, 1910, p. 15). The essential requirement for taking up banking at the time was
4259 Ch03 4/12/02 9:44 am Page 81
San Francisco Clearinghouse 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
81
apparently to be the owner of a safe in a well-protected building (Wright, 1910, p. 15). Express companies (including Wells Fargo) also collected the gold at the point of production in exchange for local or Eastern bank drafts (Wright, 1910, pp. 15–16). Between 1849–1856, private mints supplied the gold coins until the U.S. Mint in San Francisco was in full operation. Gold coin was the primary medium of exchange in California, and paper money was viewed with suspicion and seldom accepted. Even after the establishment of the National Banking System in 1864, there were initially no national banks established in California. On July 1, 1864, the Bank of California was organized as a joint stock commercial bank (the second of its kind in the U.S.) and reorganized as a national bank shortly thereafter. At the same time, the private bank of Donohoe, Kelly, and Co. was formed (Wright, 1910, pp. 19–20). In 1870, the U.S. Congress authorized the establishment of National Gold Banks that issued currency redeemable in gold. The First National Gold Bank of San Francisco was organized in November 1870 and began issuing banknotes in March 1871 (Wright, 1910, p. 51). Though the notes were issued at face value, brokers across the street from the bank were discounting the notes at an average of 90 cents on the dollar in 1871 (Wright, 1910, p. 88). It was not until January 1, 1879, when Congress resumed convertibility of greenbacks into gold, that paper money circulated at par in California. The first incorporated savings bank in California was the Savings and Loan Society of San Francisco, organized in July 1857 (Wright, 1910, p. 28). The first president of the bank was E. W. Burr, who at the time was the mayor of San Francisco; he retained his position with the bank until 1880. Other savings banks that came into existence around this time were the Hibernia Savings and Loan Society, French Savings and Loan Society, the California Building Loan and Savings Society, and the San Francisco Savings Union (Wright, 1910, p. 32). By 1906, there were 14 savings banks in San Francisco (Wright, 1910, p. 38). Though savings banks had been providing annual financial statements for some time, chartered banks were not required to prepare statements until July 1876. The private banks generally ignored the law (Wright, 1910, p. 75). The Board of Bank Commissioners was authorized by the legislation to begin regular bank examinations and they did so starting in the fall of 1878 (Wright, 1910, p. 108). Of the first five banks examined (all savings banks in San Francisco), four were closed shortly after the examinations. These were followed by voluntary liquidations of at least a half dozen interior savings banks (Wright, 1910, pp. 108–109). There was a net loss of eight savings banks and a net gain of two commercial banks in the first two years of the operation of the new law (Wright, 1910, p. 135). In 1887, legislation required private banks to submit 81
4259 Ch03 4/12/02 9:44 am Page 82
82 1 2 3
RONNIE J. PHILLIPS
semi-annual reports on the same dates as the incorporated banks, but it was not until 1905 that the private banks were placed under the jurisdiction of the Board of Commissioners (Wright, 1910, p. 136).
4 5
The Establishment of the San Francisco Clearinghouse
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
The establishment of the San Francisco Clearinghouse (SFCH) was in part a response to demands by the public for more disclosure of bank conditions in the wake of bank failures in 1875. The SFCH was established on February 4, 1876 with fifteen banks as members. Payments at the SFCH were made in gold coin unless checks were specifically payable in silver or currency (Chandler, 1986, p. 6). On June 1, 1883, the Clearinghouse began using U.S. Treasury Gold Certificates for settlements, though there was a temporary return to gold coin after the Panic of 1893 and certificates were not used again until August 19, 1899. By 1902, 70% of the settlements were in Treasury Gold Certificates (Chandler, 1986, p. 7). During the Panic of 1893, the clearinghouse permitted banks to settle balances with clearinghouse loan certificates. Table 1 indicates the growth of banking in San Francisco over the period 1858–1908. The impact of the Panic of 1907 and the national economic downturn is presumed to account for the decline in total deposits in 1908. Table 2 lists the founding members of the SFCH and the type of charter. At the time, three types of charters were possible: state, national, or foreign. In addition, banks could operate without a charter as private institutions. The growth in payments activity of the SFCH is indicated in Table 3. The decline in 1893 can be attributable to the panic in that year, and the decline in 1908 to the Panic of 1907. Table 1.
29
Deposits of Chartered San Francisco Banks, Selected Years, 1858–1908 (thousands of dollars).
30 31 32 33 34 35 36 37 38 39 40
Year 1858 1868 1878 1888 1898 1906 1907 1908
Savings Banks $20 $17,166 $60,631 $63,154 $105,192 $169,538 $160,965 $147,095
Source: Wright, 1910, pp. 69–71.
State chartered commercial banks ––– ––– $39,806 $25,526 $28,659 $80,875 $101,902 $70,225
Nationally chartered banks
Total
––– ––– $1,331 $3,048 $14,916 $36,954 $52,135 $31,785
$20 $17,166 $101,768.00 $91,728.00 $148,767.00 $287,367.00 $315,002.00 $249,105.00
4259 Ch03 4/12/02 9:44 am Page 83
San Francisco Clearinghouse
Table 2.
1 2
83
Founding Members of the San Francisco Clearinghouse, March 1876.
3 4 5 6 7 8 9 10 11 12 13 14 15
Bank
Type of Charter
The Anglo-Californian Bank, Ltd. The Bank of British Columbia The Bank of British North America The Bank of San Francisco B. Davidson & Co. Ir. Belloc Donohoe, Kelly & Co. Hickox & Spear London and San Francisco Bank, Ltd. The Merchants’ Exchange Bank Sather & Co. Swiss-American Bank Wells Fargo & Co.
Foreign (1873) Foreign (1840) Foreign State Private Private Private (1864) Private Foreign State Private Foreign Private (1852)
16 17
Source: Records of the San Francisco Clearinghouse.
18 19
Table 3.
20 21 22 23 24 25 26 27 28 29 30 31
Annual Clearings of the San Francisco Clearinghouse, 1876–1908, Selected Years.
Year
Clearings
1876 1878 1888 1893 1898 1906 1907 1908
$476,123,238 $715,329,320 $836,735,954 $699,285,778 $813,153,024 $1,998,400,779 $2,133,883,626 $1,757,141,850
Source: Wright 1910, pg. 123.
32 33 34 35 36 37 38 39 40
In 1905, the California legislature passed a provision in the banking laws establishing minimum levels of capital for banks graded by the population of the community. This provision was ruled unconstitutional and was repealed in 1907 (Wright, 1910, p. 141). Instead, a provision was passed that required that the capital should be at least 10% of deposits, with an upper limit of paid-up capital of $1,000,000, and no bank could be organized with less than $25,000 capital (Wright, 1910, p. 141). 83
4259 Ch03 4/12/02 9:44 am Page 84
84 1
RONNIE J. PHILLIPS
III. THE ROLE OF CLEARINGHOUSES
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
Though government involvement in the payments system has been regarded as the norm, the theory and practice of self-regulated payments systems has been a topic of recent research (Gorton, 1989; Gorton & Mullineaux, 1987; Selgin & White, 1994). In their study of the Suffolk banking system, Charles Calomiris and Charles Kahn (1996) present the arguments in favor of private sector payments systems vs. government mandated systems. The potential benefits, they note, are as a means to provide stability during periods of financial stress and as a means to clear payments during normal times (Calomiris & Kahn, 1996, p. 772; McAndrews & Roberds, 1995). Because cooperative systems involve coinsurance, they reduce the likelihood and costs of financial crises. In addition, interbank coordination reduces transactions and information costs during normal times. The cooperative arrangements reduce the costs of using bank liabilities as a means of payment and increase the demand for banknotes. Bankers, because of the information specialization that arises with financial intermediation, are able to determine the optimal allocation of investment funds at their disposal. However, an incentive problem arises between the banker and the depositors. As Calomiris and Kahn (1991) have shown, a demandable debt contract with a “first come, first served” rule can provide a solution to this problem (Calomiris & Kahn, 1996, p. 773). A cooperative arrangement whereby one bank places deposits in another bank helps the standardization of products, services, and practices. It also provides an enhanced mechanism for its liability holders to run the bank. Further, private regulation reduces the costs of monitoring one another, and membership in a cooperative arrangement provides a credible signal to uniformed depositors about a bank’s quality (Calomiris & Kahn, 1996, p. 773). Though government intervention could provide these advantages, with private sector solutions the costs are paid by the private sector. Because of incentive problems of government regulators, it may also be the case that private arrangements may be able to act more quickly in preventing fraud and other problems that could threaten the entire arrangement (Calomiris & Kahn, 1996, p. 774). Clearinghouses were private sector solutions that arose in the nineteenth century and served three important functions: (1) settlement of payments; (2) regulatory body; and (3) lender of last resort (U.S. National Monetary Commission, 1910a). These three functions incorporated a mix of competitive and anti-competitive actions, and, consequently were encumbered with mixed costs and benefits. The first and foremost role of a clearinghouse was the multilateral netting of payments. As stated in Article II of the Constitution of the San Francisco Clearinghouse:
4259 Ch03 4/12/02 9:44 am Page 85
San Francisco Clearinghouse 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
85
The objects of this Association shall be the effecting at one place of the daily exchanges between the several associated Banks, and the payment at the same place of the balances resulting from such exchanges.
The clearinghouse was responsible only for those balances actually paid into the hands of the clearinghouse manager. In the event of losses of balances while in the hands of the manager, for which he could not be held responsible, these were borne and paid by the members of the association in an equal proportion. Clearly, to the extent that the settlement of payments is the primary function of the clearinghouse, then the agreement to meet at a central location, and at a certain time, enables all banks to minimize their costs of settlement. The need for personnel to make individual settlements with each bank is reduced. As long as there is never a question of final settlement, then there is no risk to the clearinghouse. There is an incentive for the clearinghouse association to allow only those institutions that are able to make settlements good with regularity. In order to reduce costs for individual settlements further, the clearinghouse would require that each member bank place on deposit with the clearinghouse a certain amount of the final means of settlement. In the 19th century this was primarily gold coin, but could also be silver coin, various types of government currency, and possibly national banknotes. A clearinghouse serving to reduce the transactions costs of payments of individual banks also reduces the overall social costs of operating the payments system. If this was the sole function of clearinghouses, then they would have undeniably served a function that was consistent with a cooperative private sector solution to the costs of operating the payments system in a competitive market economy. To this extent, the clearinghouses were beneficial to the overall economy. However, clearinghouses did not limit their activities to merely reducing the social costs of the payments system. Rather, clearinghouses were a means to regulate competition in the industry since clearinghouse membership in some sense represented a “seal of approval” for banking firms from a body that represented the most prominent banks in the city. Clearinghouses went further in many instances by imposing minimum fees, interest rate ceilings, reserve requirements, examinations, and so on, which were clearly attempts to restrain competition. In this respect, the actions of clearinghouses may not have been consistent with the promotion of a competitive market economy. After all, in a market economy, others besides banks would be better at assessing the viability and performance of a member bank. The final function of a clearinghouse was as a lender of last resort during a financial panic. It is this function, together with the regulatory function, that Gorton and Mullineaux have in mind when they refer to the clearinghouse’s 85
4259 Ch03 4/12/02 9:44 am Page 86
86 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
RONNIE J. PHILLIPS
role as the “joint production of confidence.” (Gorton & Mullineaux, 1987). The fundamental question is why firms that are in competition with each other would voluntarily enter into the joint production of confidence. For Gorton and Mullineaux, the answer lies in the problem of the distribution of asymmetric information in the banking industry. The problem lay in the mix of demand deposits and banknotes that came to predominate in the late nineteenth century. Gorton and Mullineaux argue that the capacity of the market to monitor and control banks eroded as demand deposits came to supplant bank notes in the nineteenth century (Gorton & Mullineaux, 1987, p. 458). Hierarchical organization and control replaced the market mechanism as Coase (1937), Williamson (1975), Chandler (1977) and Stiglitz (1985) have argued. In periods of financial panic, the clearinghouses began to issue clearinghouse loan certificates (CHLC). Because each member bank had made a contribution in the final means of payment (gold for the most part), the clearinghouse could effectively pool these reserves and loan them to members experiencing problems. This action brings attention to the fundamental contradiction in the clearinghouse arrangement. Banks that were not experiencing problems had no incentive to loan to their competitors to keep them in business. Indeed, the problems of competitors should have been welcomed since it would have meant that depositors would shift to the banks viewed as better operated and managed. However, there is the problem of contagion – a run on one bank leading to runs on other banks. This would arise if depositors did not have complete information about the solvency of other banks. Hence, solvent banks might have had an incentive to help out banks in trouble if they feared a contagion. The fear of contagion provided the rationale for the clearinghouse to lend the pooled reserves to a member in trouble and it did this through the issuing of clearinghouse loan certificates. The bank in trouble pledged collateral (securities of some sort) and the clearinghouse advanced a discounted amount of clearinghouse loan certificates. The bank paid out the CHLC to depositors who wished to withdraw. The holder of the CHLC could then, if he/she wished, present the CHLC for payment in gold to the clearinghouse. Alternatively, the CHLC could function like any other paper currency. However, the issuing of CHLC to be used as currency was not legal until the passage of the AldrichVreeland Act in 1908. Numerous studies have been made of the issuing of CHLC and in general the conclusion is that the system worked sometimes and at other times it did not. The role of the Treasury was often important in the panics (Timberlake, 1993; Dwyer & Gilbert, 1989). The Federal Reserve Act provided the established clearinghouses with competition from the Fed – which was able to subsidize check clearing, unlike the existing clearinghouses (Spahr, 1926;
4259 Ch03 4/12/02 9:44 am Page 87
San Francisco Clearinghouse 1 2 3 4 5 6 7 8 9
87
Stevens, 1996; Stevens, 1998; Gilbert, 2000). In summary, the development of clearinghouses represented an economic organization formed by economic agents (bankers) in order to counter the impact of unrestricted competition. As Gorton notes with regard to clearinghouses: That such an economic entity should have endogenously arisen in the banking industry suggests important links between the characteristics of the product and institutional and contractual forms of economic organization. While much work remains to be done on these links, the existence of the clearinghouse suggests that private agents can creatively respond to market failure (Gorton, 1985, p. 283).
10 11 12
IV. THE EARTHQUAKE OF 1906
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
At 5:12am on Wednesday, April 18, 1906, an earthquake with a magnitude of 8.3–8.6 on the Richter scale struck San Francisco. The earthquake and fires that burned until Saturday, the 21st, destroyed 500 city blocks, 25,000 buildings, left 250,000 homeless, killed 450–700, with total damages in 1906 dollars at between $350 and $500 million. The destruction included the financial district of San Francisco and most bank buildings (Thomas & Witts, 1971). By early in the morning, San Francisco Mayor Eugene Schmitz had issued a proclamation in which he announced that federal troops, policemen, and special officers had the authority to kill any persons found looting or involved in other criminal activity. All gas and electricity was cut off indefinitely (to avoid spreading fires), all citizens were to remain in their homes, and the Mayor appointed a Citizen’s Relief Committee of Fifty (Muscatine, 1975, p. 433). Much has been written on events of the earthquake, some of which is pertinent to the impact on banking and the payments system. The earthquake damage itself was extensive, but the fact that water mains were broken and fires had been started from busted gas mains, meant that those who survived the earthquake faced the prospect of losing their homes and businesses to fires. In addition, some fires were intentionally set and some buildings dynamited in order to help contain the fires. Thus, the earthquake did not level the banks for the most part, but the subsequent fires destroyed, or heavily damaged, the buildings of virtually all of the banks in the city. The fact that the initial earthquake, though creating panic and chaos in the city, did not destroy all the banks (or even very many), meant that one could have hours or days in which to make plans. One banker who was neither a member of the Mayor’s Committee of Fifty nor the SFCH, but who did respond quickly was A. P. Giannini. 87
4259 Ch03 4/12/02 9:44 am Page 88
88 1
RONNIE J. PHILLIPS
A. P. Giannini and the Bank of Italy
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
When his father-in-law passed away in 1902, A. P. Giannini, who had been a successful entrepreneur and businessman, assumed his position as a director of a bank, the Columbus Savings and Loan Society, a small bank catering almost exclusively to Italian-Americans in the community of North Beach in San Francisco (Nash, 1992, pp. 22–23). After disagreements with other members of the Board of Directors, Giannini established his own bank, the Bank of Italy, in 1904. When the earthquake struck, the Bank of Italy was a small but growing and successful bank. Giannini was not in San Francisco when the earthquake struck, but managed to arrive shortly thereafter. He realized that, even though the initial earthquake had not destroyed his bank building, it was quite likely that it would be consumed by fire within a short period of time. When Giannini arrived at the bank around noon, he found that his Assistant Cashier and clerk were standing at their desks. Both men had arrived at the bank early in the morning and, finding the building intact, decided to open for business as usual. They went to the Crocker Woolworth National Bank where they collected three heavy canvas bags which held over $80,000 in gold and silver coins – every cent of hard cash held by the Bank of Italy (Thomas & Witts, 1971, p. 118). They then went back to the bank and opened for business, though by 11 AM they were apparently questioning the wisdom of doing so. When Giannini arrived he reportedly told them “only an insane man would open a bank on a day like this. You should have left the money at Crocker. At least it would have been safe” (Thomas & Witts, 1971, p. 120). By 5 p.m. the day of the earthquake, Giannini and his staff loaded the bank’s available cash and $80,000 in gold, along with bank records, onto a wagon and proceeded to make their way to Giannini’s home in San Mateo. Before departing, he parked his wagons at his brother-in-law’s home near Francisco and Jones streets at the far end of the North Beach Area (Nash, 1992, p. 32). Around 8 PM, Giannini began his trip to San Mateo, and arrived there – amid great panic and chaos – in the early morning hours of April 19. If Giannini could get his cash out, what did the other bankers do with the time between the earthquake and fire? At the Crocker Bank on the morning of the 18th, customers brought ledgers, records, and money for storage in the vaults. However, William Crocker, the bank’s president, had already hired a boat to take all the bank records out into the center of the Bay – “and just sit there” (Thomas & Witts, 1971, p. 119). At the Anglo-California Bank, tellers carted negotiable bonds in a wheelbarrow down Market Street to the Ferry building for shipment across the Bay to Oakland. In three hours they shifted a
4259 Ch03 4/12/02 9:44 am Page 89
San Francisco Clearinghouse 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
89
million dollars worth of unregistered bonds (Thomas & Witts, 1971, p. 119). At the London, Paris, and American Bank, sacks of U.S. registered mail were placed in the vault along with deposits of jewel boxes and family deeds (Thomas & Witts, 1971, p. 119). On Friday the 20th, Giannini returned to San Francisco where he learned that the Bank of Italy building had indeed been burned to the ground. The bank’s safe had become a lump of molten iron. On Saturday, April 21st, Giannini attended a meeting of businessmen near the waterfront. In response to the view that the banks would not be able to open until November, he responded: Gentlemen, to follow the course you are suggesting will be a vital mistake . . .. We’ve got to fight our way out of this spot. If you keep your banks closed until November . . . there will be no city or people left to serve. Today is the time they need you. The time for doing business is right now. Tomorrow morning I am putting a desk on Washington Street wharf with a Bank of Italy sign over it. Any man who wants to rebuild San Francisco can come there and get as much cash as he needs to do it (quoted in Nash, 1992, p. 33).
On Sunday, the 22nd, just four days after the earthquake, Giannini opened for business, bringing $10,000 in cash from his San Mateo hoard. He took out an ad in the San Francisco Chronicle and announced that he was open for business (Nash, 1992, p. 34). Giannini actively began to make loans, and his actions after the earthquake made him something of a local hero, and of course were great publicity for his bank. The actions of the SFCH can be contrasted with those of an individual banker who was not bound by the agreements reached by the clearinghouse members. Indeed, Giannini was viewed as an outsider, “someone to be watched, possibly, but never to be admitted to that intimate circle dominated by the American National Bank and Wells Fargo” (Thomas & Witts, 1971, p. 43). Giannini’s bold decision to reopen so quickly boosted his image in the community and aided the growth of his bank in later years. The earthquake also provided the rationale for branch banking, an area where Giannini was very aggressive. Some bankers removed their gold and cash while others left everything in the vaults. Removing the gold and cash was an enormous risk both from the threat of robbery and the danger that it would provoke a further banking panic. Giannini was bold and lucky and reaped the benefits in subsequent years as his bank grew in assets and reputation.
36 37
The Saving of the U.S. Mint
38 39 40
On the morning of the 18th, rumors spread that an armed gang was massing to storm the U.S. Mint, which contained $300 million in gold coin and bullion. 89
4259 Ch03 4/12/02 9:44 am Page 90
90 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
RONNIE J. PHILLIPS
This was indeed a rumor, but in any event a company of troops from the Sixth Infantry took up defensive positions on the roof that gave them an excellent vantage point for firing upon intruders (Thomas & Witts, 1971, p. 127). The Mint in San Francisco was established in April 1854, and the building at Fifth and Mission was a monolith of granite and sandstone blocks and suffered only minor damage in the earthquake. However, as the fire approached, the troops closed and bolted the heavy iron shutters across the ground-floor windows. The tarred roof was ripped away with iron bars and pick shovels, then buckets of blue vitriol were hoisted up from the refinery and mopped over the exposed roof beams. Though the Mint had its own water supply, the quake had broken the pump; but it was repaired so that it could be hand-operated. The Mint was saved during the fire due to its construction, the distance of the façade from the street, and an artesian well on the premises, and employees who helped fight the fire (Wright, 1910, pp. 164–165). The Sub-Treasury building on Commercial street, which had at one time been occupied by the Mint, was destroyed in the 1906 earthquake and fire. The contents of the vault were not damaged, however (Wright, 1910, p. 169).
18 19
The Response of the San Francisco Clearinghouse(1)
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
There was a brief meeting of the SFCH at 9 AM on the day of the earthquake where it was decided that the banks should remain closed. By the evening of that day, virtually all of the bank buildings had been destroyed. This meeting was followed by two brief meetings on