The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
Financial Panics, the Seasonality of the Nominal Interest Rate, and the Founding of the Fed By JEFFREY A. MIRON* After the founding of the Fed in 1914,1 the frequency of financial panics and the size of the seasonal movements in nominal interest rates both declined substantially. Since the Fed was established in part to "furnish an elastic currency,"2 it is natural to hypothesize that the Fed caused these changes in the behavior offinancialmarkets. There were, however, a number of other major changes in the economy and in the financial system during this period including World War I, the shift from agriculture to manufacturing,3 and the loosening of the gold standard.4 Moreover, Robert Shiller (1980) has examined the effect of the Fed's founding on the seasonal in real interest rates and has concluded that the Fed's actions had little or no effect. This paper investigates the relationship between financial panics, seasonal movements in nominal interest rates, and the open market operations of the Fed after 1914. The paper establishes that the Fed, by carrying out the seasonal open market policy that eliminated the seasonal in nominal interest rates, caused the decrease in the frequency of panics. Since seasonal movements are anticipated and financial panics are probably real events, the results show that an anticipated monetary policy had real effects on the economy. The issue of whether anticipated monetary policy can affect real variables, which is at the heart of monetary economics, has received much recent attention following the well-known contributions by Robert Barro (1977, 1978). His results have been subjected to a barrage of critical review, much of it supporting his finding that only unanticipated changes in money have real effects (for example, Barro and Mark Rush, 1980; Robert Litterman and Lawrence Weiss, 1985; Robert Lucas, 1973; Shiller; Christopher Sims, 1980),5 some of it arguing that the evidence rejects the neutrality of anticipated money (for example, Robert Gordon, 1982; Frederic Mishkin, 1982, 1983).6 The generally inconclusive nature of the debate reflects the difficulty of determining whether policy caused or responded to changes in the economy and of distinguishing anticipated from unanticipated policy actions. Thomas Sargent (1976), when describing the possible observational equivalence of classical and nonclassical models, suggested that identification would be aided if it were possible to draw data from two different policy regimes. * Department of Economics, University of Michigan, Ann Arbor, MI 48109. I thank Stanley Fischer, Larry Summers, Peter Temin, Olivier Blanchard, Milton Friedman, Steve Zeldes, Steve O'Connell, Sue Collins, Robert Clower, and two anonymous referees for helpful comments on earlier drafts of this paper. 1 The Federal Reserve Act was passed by Congress on December 23, 1913. The Board of Governors took office and began planning the organization of the System on August 10, 1914. The twelve banks opened for business on November 16, 1914. 2 This quote is from the preamble to the Federal Reserve Act. 3 The share of agriculture in Gross Domestic Product fell from 24 percent in the period 1897-1901 to 12 percent in 1922. See Historical Statistics of the United States,... (1976, Series F125-129, p. 232). 4 During World War I, several countries (including Great Britain) left the gold standard, so the United States was less affected by external conditions. 5 The approach to testing neutrality in Barro and Rush is the same as in Barro (1977, 1978). LittermanWeiss and Sims use vector autoregressive techniques and base their conclusions on the failure of money to be Granger causally prior for real income. Lucas shows in a cross section of countries that the variance of money shocks is negatively correlated with the variance of output movements. 6 Barro (1978) introduced the use of cross-equation restrictions into this literature. This more powerful way of testing neutrality has been exploited extensively by Mishkin (1982, 1983), who has usually found that the data reject neutrality, contrary to the results of Barro. 125 VOL. 76 NO. 1 MIRON: FOUNDING OF THE FED the loans made by private banks should have become less seasonal. Section II examines empirically the implications of the model and the hypotheses it suggests about the behavior of the Fed. These results come from a simple model, but they do not depend on the particular assumptions made in order to keep the analysis simple.11 The model presented above is the most complicated one that can be tested empirically; it is not possible to test the additional implications of more complicated models because of data limitations. II. The Evidence A. Historical Background: The National Banking System and the Founding of the Fed The period from 1863 through 1913 is known as the period of the National Banking System because the provisions of the National Banking Acts of 1863, 1864, and 1865 determined the banking and financial structure in several critical ways. The National Banking Acts were both a response to problems of the financial system that existed before the Civil War and a measure designed to raise revenue for the North during the war. The Acts successfully generated revenue and cured some prewar financial ills (notably the multiplicity of note issue). During the National Banking Period, however, those in academia, the banking community, and government still regarded the financial system as fundamentally flawed because of the "perverse elasticity of the money supply" and the high frequency of financial panics. The term perverse elasticity of the money supply referred to the tendency of the money supply to contract in precisely those periods when it was "needed" most. This occurred in the spring and fall of each year when seasonal increases in loan and currency demand forced interest rates up and reserve-deposit ratios down. These seasonal movements in loan and currency demand were attributed 11 AppendicesA and B to ch. IV of my disseration show that the conclusions are still valid if one allows for a pyramided banking system, or for the general equilibrium interactions of the economy. 129 mainly to the need for both currency and credit by the agricultural sector of the economy in the spring planting season and the fall crop-moving season, and to the need for currency and credit by the corporate sector for quarterly interest and dividend settlements. Additional currency was needed because the volume of transactions was higher in these periods. Credit demand was high because farmers borrowed to finance the planting and harvesting of the crops.12 The financial panics that occurred in this period were combinations of bank failures, bank runs, and stock market crashes. A typical panic began after an individual bank was hit by either an unexpectedly large deposit withdrawal or a large loan default. If the bank had a small amount of reserves, it would need to call in some of its loans. This might concern other banks enough so that they would call in some of their loans, many of which were in stock market call loans, and the cumulative effect of loan recall by many banks tended to depress the stock market. At the same time, the fact that banks were calling in loans caused the nonbank public to increase its desired currency-deposit ratio, and this could cause either individual bank failures or runs on many banks. Eventually the process either reversed itself or ended in a suspension of convertibility.13 There were, of course, differences in the dynamics of various panics. Some began in New York as the result of a large loan default at a New York bank and then were transmitted West as New York banks tried to acquire additional reserves from the country banks. Others started in the West when crop failures damaged the liquidity positions of country banks who then tried to recall 12 E. W. Kemmerer (1910, pp. 223-24) mentions increased rail and barge activity during warm weather and holiday seasons as additional reasons for seasonal activity in the financial markets. A. Piatt Andrew (1906) discusses the influence of agriculture on economic activity during the pre-Fed period, and J. Laurence Laughlin (1912, pp. 309-42), discusses the seasonal cycle in general economic activity. See also O. M. W. Sprague (1910), C. A. E. Goodhart (1969), and John James (1978, pp. 127-37) for discussions of the seasonal flows within the country that accompanied the seasonal changes in interest rates and reserve positions of banks. 13 Sprague (pp. 1-225). 130 THE AMERICAN ECONOMIC REVIEW balances from reserve cities. Nevertheless, the key element of a panic was the same in all of the major episodes. This key element was a generally increased demand for reserves that could not be satisfied for all parties simultaneously in the short run. The likelihood that an event such as a large loan default would precipitate a panic depended on the initial position of the banking system. If such an event happened at a time when loan demand was high or deposit demand was low, so that the reserve-deposit ratios of banks were low, then the costs imposed by the loan default were higher. Since there were seasonal movements in loan and deposit demands that produced seasonal movements in reserve-deposit ratios, panics tended to occur in the fall and spring, when high-loan demand and low-deposit demand produced low reserve-deposit ratios. Thus the problems of perverse elasticity and the accompanying financial panics were partly a result of and coincided with the seasonal movements in asset demands. The academics, bankers, and government officials of the time understood this phenomenon. J. Laurence Laughlin, a professor of economics at the University of Chicago, commented in detail on this relation between panics and seasonality in his 1912 treatise on reform of the banking system (pp. 309-42). Paul Warburg, a Wall Street banker who later served on the Federal Reserve Board, wrote in 1910 that "there can be no doubt whatever that the basis for healthy control by a central bank must exist in a country where regular seasonal requirements cause, with almost absolute regularity, acute increased demand for money and accommodation" (1930, p. 156). Leslie Shaw, Secretary of the Treasury from 1902 to 1906, actively attempted to accommodate the seasonal demands in financial markets, although the funds available to him were not sufficient to allow him to be successful.14 The panic of 1907 precipitated sufficient concern about panics and elasticity that 14 See Andrew (1907, p. 559), and Richard Timberlake (1978, p. 181). See Andrew (1907) also for an interesting analysis of Shaw's other activities and Timberlake (1963) for a critique of Andrew's analysis. MARCH 1986 Congress passed the Aldrich-Vreeland Act of 1908. This Act addressed the problems of the banking system by granting certain emergency powers to New York City banks and by creating the National Monetary Commission. This Commission was assigned to undertake a detailed study of the U.S. banking system. Its Report, published in 1910, contained in depth examinations of every aspect of banking theory and practice in the United States and abroad. Two parts of the Report deserve particular notice. O. M. W. Sprague, a professor of economics at Harvard, wrote History of Crises Under the National Banking System. This book examined in detail the operation of the banking system during five of the worst financial crises (1873, 1884, 1890, 1893, and 1907). Sprague wrote that "with few exceptions all our crises, panics, and periods of less severe monetary stringency have occurred in the autumn" (p. 157). E. W. Kemmerer of Cornell contributed the volume Seasonal Variations in the Relative Demands for Money and Capital in the United States. He noted that "the evidence accordingly points to a tendency for the panics to occur during the seasons normally characterized by a stringent money market" (p. 232). Thus two parts of the Report mentioned explicitly the tendency for panics to occur in certain seasons of the year. The Federal Reserve Act established the Federal Reserve System in 1913, three years after the publication of the Commission's Report. The preamble to the Act states that it is "an act to... furnish an elastic currency." It was to be expected, therefore, that the Fed would try to eliminate panics by accommodating the seasonal demands in financial markets. B. Evidence of the Changes in Financial Markets I now document the two facts cited in the introduction: the frequency of financial panics diminished after the founding of the Fed; and the size of the seasonal fluctuations in nominal interest rates diminished also. Table 1 shows the starting dates of the financial panics that occurred during the period 1890-1908 according to Sprague and TABLE 1—STARTING DATES AND CLASSIFICATION OF FINANCIAL PANICS ACCORDING TO SPRAGUE AND KEMMERER Classification Sprague Financial Stringency Crisis Crisis Kemmerer Major Panics Minor Panics Year Month 1890 1893 1907 August May October 1890 1893 1899 1901 1903 1907 1893 1895 1896 1896 1898 1899 1901 1901 1902 1904 1905 1906 1906 1907 1908 September May December May March October February September June December March September July September September December April April December March September Sample Period Dependent Variable Significance F- Statistics Level 1890-1908 vs. 1919-28 Nominal Interest Rate Loans-Reserve Ratio Loans Reserves Sample Period Significance F-Statistics Level Dependent Variable 1890-1908 Nominal Interest Loans-Reserve Loans Reserves 1919-28 Nominal Interest Loans-Reserve Loans Reserves 1922-28 Reserve Credit Total Credit Rate Ratio Rate Ratio 1.68 4.28 2.46 4.90 2.05 4.90 3.65 1.90 7.09 5.54 .003 .000 .000 .000 .000 .000 .000 .000 .000 .000 2.05 4.90 3.65 1.90 .000 .000 .000 .000 January February March April May June July August September October November December 0 1 3 2 2 1 1 0 6 1 0 4 VOL. 76 NO. 1 MIRON: FOUNDING OF THE FED (1929, 1930, 1932) while two were in the spring (1931, 1933).20 Thus the recurrence of panics during this period corroborates the hypothesis that the Fed caused the reduction in the frequency of panics after 1914. The decreased accommodation of the Seasonals in asset demands was probably the result of a generally restrictive open market policy that the Fed initiated in late 1928.21 During much of the 1920's, there was fear that speculation by participants in the stock market was "excessive," and those who objected to the speculation most encouraged the Fed to restrain the growth of credit, particularly loans by banks to stock market brokers. The officials at the Fed differed in their view of how much to restrain credit. On balance, however, they opposed restraining speculation so much that it might adversely affect general business activity. This policy changed toward the end of 1928. Stock market speculation had been especially virulent, and the Fed responded with a strongly restrictive policy. The explanation for the change in policy is that Benjamin Strong, the governor of the New York Fed, died in October of 1928.22 During the period 1915-28, Strong was a dominant force in the Federal Reserve System and in the entire financial community. In the words of his biographer, Lester Chandler, Strong was "one of the world's most influential leaders in the fields of money and finance. During the first fourteen turbulent, formative years of the Federal Reserve System, his was the greatest influence on American monetary and banking policies" (1958, p. 3). Strong intensely disliked stock market speculation, but was an outspoken critic of restraining speculation at the cost of causing a recession. His death allowed the balance of opinion at the Fed to shift toward greater restraint, and a highly restrictive policy resulted. One of 20 See F r i e d m a n a n d Schwartz (pp. 305, 308, 313, 324). 21 See Paul Trescott (1982) for a more detailed examination of this aspect of F e d policy. 22 F r i e d m a n a n d Schwartz (pp. 4 1 1 - 1 9 a n d p p . 6 9 2 - 9 3 ) discuss in detail the effects of Strong's death o n the power structure of the Fed. 137 the manifestations of this policy was the incomplete accommodation of the seasonal demands in financial markets. F. The Real Effects of Financial Panics The final issue discussed in this section is whether financial panics had real effects on the economy. It is not possible to test an explicit model of the real effects of panics because of data limitations.23 It is possible, however, to demonstrate support for the proposition that panics had real effects if one is willing to make assumptions about what these effects might have been. I assume here that panics affected the distribution of output by decreasing the average level of real activity, increasing the variance of real activity, and increasing the length of business cycles.24 In the context of this paper there are two implications of the proposition that panics were real events. First, the distribution of output in the pre-Fed period should have been worse in panic years than in nonpanic years; second, the distribution of output post-Fed should have been better than that pre-Fed. Table 5 shows the mean and variance of the rate of growth of annual real GNP for the period 1890-1908 and for this period minus the years in which panics occurred (Kemmerer's major panic definition). The average level of GNP growth is higher and the variance of real growth lower for nonpanic years, so these facts support the hypothesis that panics altered the distribution of output. They do not, of course, prove that hypothesis, since panics might be the result of negative output shocks. Nevertheless, the facts in the table lend plausibility to the proposition that panics changed the distribution of output during the pre-Fed period. 23 The data in Gordon are interpolations. See B e n Bernanke (1983) for a n analysis of the effects of t h e financial crises during the Great Depression, C a g a n for a discussion of the real effects of panics d u r i n g the p r e - F e d period, a n d G o r t o n (1983) for work o n t h e general relation between panics a n d business cycles. 24 Peak Trough July 1890 May 1891 January 1893 June 1894 December 1895 June 1897 June 1899 December 1900 September 1902 August 1904 May 1907 June 1908 August 1918 March 1919 January 1920 July 1921 May 1923 July 1924 October 1926 November 1927 Average for Pre-Fed Period = 17.5 months Average for Post-Fed Period = 14.25 months Length 11 18 19 19 24 14 9 19 15 14 MIRON: FOUNDING OF THE FED VOL. 76 NO. 1 above does not necessarily imply that continued elimination of interest rate Seasonals is desirable. The analysis does show that an important aspect of Fed policy is its seasonal behavior, and it demonstrates that this aspect of policy can have substantial real effects on the economy. REFERENCES Andrew, A. Piatt, "The Influence of Crops Upon Business in America," Quarterly Journal of Economics, May 1906, 20, 323-53. , "The Treasury and the Banks Under Secretary Shaw," Quarterly Journal of Economics, August 1907, 21, 519-68. Barro, Robert J., "Unanticipated Money Growth and Unemployment in the United States," American Economic Review, March 1977, 67, 101-15. , "Unanticipated Money, Output, and the Price Level in the United States," Journal of Political Economy, August 1978, 86, 549-80. and Rush, Mark, "Unanticipated Money and Economic Activity," in Stanley Fischer, ed., Rational Expectations and Economic Policy, Chicago: University of Chicago Press, 1980, 27-74. Bernanke, Ben S., "Nonmonetary Effects of the Financial Crisis in the Propagation of the Great Depression," American Economic Review, June 1983, 73, 257-76. Bryant, John, "A Model of Reserves, Bank Runs, and Deposit Insurance," Journal of Banking and Finance, December 1980, 4, 335-44. Cagan, Phillip, Determinants and Effects of Changes in the Stock of Money, 18751960, New York: Columbia University Press, 1965. Chandler, Lester, Benjamin Strong, Central Banker, Washington: The Brookings Institution, 1958. Clark, Truman, "Interest Rate Seasonals and the Federal Reserve System," unpublished manuscript, University of Southern California School of Business Administration, 1983. Diamond, Douglas W. and Dybvig, Philip H., "Bank Runs, Deposit Insurance, and 139 Liquidity," Journal of Political Economy, June 1983, 91, 401-19. Friedman, Milton, A Program for Monetary Stability, New York: Fordham University Press, 1959. , "Monetary Policy: Theory and Practice," Journal of Money, Credit and Banking, February 1982, 14, 98-118. Friedman, Milton and Schwartz, Anna J., A Monetary History of the United States, 1867-1960, Princeton: Princeton University Press, 1963. Glass, Carter, An Adventure in Constructive Finance, New York: Doubleday, Page, and Co., 1927. Goodhart, C. A. E., The New York Money Market and the Finance of Trade, 1900-1913, Cambridge: Harvard University Press, 1969. Gordon, Robert J., "Price Inertia and Policy Ineffectiveness in the United States, 1890-1980," Journal of Political Economy, December 1982, 90, 1087-117. Gorton, Gary, "Demand Deposits and Banking Panics," manuscript, Wharton School, University of Pennsylvania, 1982. , "Banking Panics and Business Cycles," manuscript, Wharton School, University of Pennsylvania, 1983. James, John A., Money and Capital in Postbellum America, Princeton: Princeton University Press, 1978. Kemmerer, E. W., Seasonal Variations in the Relative Demand for Money and Capital in the United States, National Monetary Commission, S.Doc.588, 61st Cong., 2d session, 1910. Laughlin, J. Laurence, Banking Reform, Chicago: National Citizens League for the Promotion of a Sound Banking System, 1912. Litterman, Robert B. and Weiss, Laurence, "Money, Real Interest Rates, and Output: A Reinterpretation of Postwar United States Data," Econometrica, January 1985, 53, 129-56. Lucas, Robert E., Jr., "Some International Evidence on Output-Inflation Tradeoffs," American Economic Review, June 1973, 63, 326-34. Miron, Jeffrey A., "The Economics of Seasonal Time Series," unpublished doctoral 140 THE AMERICAN ECONOMIC REVIEW dissertation, MIT, 1984. Mishkin, Frederic S., "Does Anticipated Policy Matter? An Econometric Investigation," Journal of Political Economy, February 1982, 90, 22-51. , "Does Anticipated Aggregate Demand Policy Matter? Further Econometric Results," American Economic Review, September 1983, 72, 788-802. Rasche, Robert, "A Review of Empirical Studies of the Money Supply Mechanism," Federal Reserve Bank of St. Louis Review, July 1972, 54, 11-19. Sargent, Thomas, "The Observational Equivalence of Natural and Unnatural Rate Theories of Macroeconomics," Journal of Political Economy, July 1976, 84, 631-40. Shiller, Robert, "Can the Fed Control Real Interest Rates?," in Stanley Fischer, ed., Rational Expectations and Economic Policy, Chicago: University of Chicago Press, 1980, 117-68. Sims, Christopher, "Comparison of Interwar and Postwar Business Cycles: Monetarism Reconsidered," American Economic Review Proceedings, May 1980, 70, 250-57. Sprague, O. M. W., History of Crises Under the MARCH 1986 National Banking System, National Monetary Commission, S.Doc.538, 61st Cong., 2d session, 1910. Timberlake, Richard H., "Mr. Shaw and His Critics: Monetary Policy in the Golden Era Reviewed," Quarterly Journal of Economics, February 1963, 77, 41-54. , The Origins of Central Banking in the United States, Cambridge: Harvard University Press, 1978. Trescott, Paul B., "Federal Reserve Policy in the Great Contraction: A Counterfactual Assessment," Explorations in Economic History, July 1982, 19, 211-20. Warburg, Paul M., The Federal Reserve System: Its Origins and Growth, Vol. II, New York: Macmillan, 1930. Willis, H. Parker, The Federal Reserve: A Study of the Banking System of the United States, New York: Doubleday, Page and Co., 1915. Board of Governors of the Federal Reserve System, Annual Report, Washington: USGPO, 1914-33. U.S. Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1970, New York: Basic Books, 1976.
@FedFRASER