¹ The quotation is from the Commercial and Financial Chronicle, July 9, 1932, p. 168.Google Scholar See Epstein, Gerald and Ferguson, Thomas, “Monetary Policy, Loan Liquidation, and Industrial Conflict: The Federal Reserve and the Open Market Operations of 1932,” this JOURNAL, 44 (12 1984), pp. 957–83;Google Scholar and Coelho, Philip R. P. and Santoni, G. J., “Regulatory Capture and the Monetary Contraction of 1932: A Comment on Epstein and Ferguson,” this JOURNAL, 51 (03. 1991).Google Scholar

² Coelho and Santoni, “Regulatory Capture,” p. 188.Google Scholar

³ See Epstein and Ferguson, “Monetary Policy,” especially pp. 968–77.Google Scholar More recently, Temin, Peter, in Lessons of the Great Depression (Cambridge, MA, 1989), has emphasized the broad role of the gold standard.Google Scholar

⁴ See Epstein and Ferguson, “Monetary Policy,” for example, p. 966.Google Scholar

⁵ The list comes from various minutes of either the Fed or the New York Fed of the period; from papers in the Ogden Mills or Eugene Meyer papers at the Library of Congress; the Herbert Hoover Presidential Library in West Branch, lowa; or papers of various private bankers cited in Epstein and Ferguson, “Monetary Policy.”Google Scholar

⁶ See Epstein and Ferguson, “Monetary Policy,” p. 961, on W. Randolph Burgess and Winfield Riefler. Strong appears never to have subscribed to the doctrine, either, though the point exceeds the scope of this reply.Google Scholar

⁷ Appendix I contains a full account of our sources of data and definitions. For the currency ratio in this period, see Friedman, Milton and Schwarz, Anna J., A Monetary History of the United States, 1867–1960 (Princeton, 1963), pp. 802–3, table B-3, and p. 333, chart 31.Google Scholar

⁸ The precise value of course depends on the time period; the appropriate time period, in turn, depends on one's view of the time horizon that policy makers would have relied on in forming their views. From Aug. 31, 1931, to Mar. 1933, both the money and credit multipliers, no matter how measured, declined about 44 percent. Because policy makers surely required some time to assimilate changes, this earlier base point is perhaps the most reasonable. Using other base points and periods yields a variety of very unstable answers: from Aug. 31, 1931, to Dec. 1932, the decline is about 31 percent; from Jan. 1932 to Mar. 1933 (the period of our critics' graph), the decline is approximately 32 percent; from Jan. to Dec. 1932, the decline is about 16 percent. The results do not change whether one calculates money or credit multipliers or varies definitions.Google Scholar

⁹ See his remarkable defense of the open market program, quoted in Commercial and Financial Chronicle, April. 30, 1932, p. 3142.Google Scholar

¹⁰ There were very few bills bought during this period because there were comparatively few to buy. The Fed, for example, could also purchase trade acceptances. But with the collapse of international trade, there were simply not enough of these available. Accordingly, the practical measure of the program became the number of government securities purchased.Google Scholar

¹¹ The question of exactly who thought what and when is too complicated to discuss here in detail. Various documents are quite clear that different advocates of reflation expected lags of various durations. Moreover, Russell Leffingwell at Morgan and others were not uniformly consistent later in recalling their own views at the time.Google Scholar

¹² Epstein and Ferguson, “Monetary Policy,” p. 976.Google Scholar

¹³ Shiller, Robert, Market Volatility (Cambridge, MA, 1989), p. 52.Google Scholar

¹⁴ It also appears that their sample of Chicago banks has some serious flaws. It excludes what even they would presumably concede was one of the classic regulatory failures of American history (in which Reconstruction Finance Corporation Chair Charles Dawes resigned to accept what wags called the “second Dawes Loan”: a gigantic bailout of his Central Republic Trust Co.—at the time probably the third largest bank in the city—from the agency he was exiting, while his stock opened at 50, fell to one, and then closed at what was almost certainly a strongly supported four bid and six asked). See, for example, James, F. Cyril, The Growth of Chicago Banks (New York, 1938), vol. 2, pp. 1037ff. Note that Coelho and Santoni's Chicago sample includes five banks, so that omission of such a large, ill-fated bank matters. There is also the fact that some bank stocks were being artificially supported during this period in both New York and Chicago, and this may cause serious problems with “efficient markets” approaches. Furthermore, another bank in their sample, Continental Illinois, had to be rescued from bankruptcy not long after the point where their analysis stopped following the stock. Added to Central Republic, this implies that two of the three largest banks in the city folded. We are also puzzled that The New York Times, which Coelho and Santoni indicated as their source for weekly stock quotations, stopped quoting prices for the stocks of the Chicago banks for two weeks during this period; we are unclear what this means.Google Scholar

¹⁵ A classic reference is Fama, Eugene et al. , “The Adjustment of Stock Prices to New Information,” International Economic Review, 10 (02. 1969), pp. 1–21.Google Scholar For review, see Schwert, William, “Using Financial Data to Measure Effects of Regulation,” Journal of Law and Economics, 24 (04 1981), pp. 121–58. Most tests focus on stock returns, which in addition to changes in stock prices, include dividends and other payments. Given Coelho and Santoni's focus on stock prices and the likelihood that over the month-long period of our test most changes in returns are accounted for by changes in stock prices, we use stock prices as an approximation.Google Scholar

¹⁶ As indicated by the significant results for Apl. 13 in the column of T-Differences. Note that the effect we are discussing concerns the announcement of the program. It is possible that a similar effect also could be detected for its termination in June. We ourselves have not run this test, because of the comparatively enormous amount of data that such tests require. (For this reason, for our tests we used a sample of seven very large New York banks, as listed in Appendix 1.) A test in June, however, might run afoul of the truly gigantic bank runs that were then occurring in Chicago and the anxieties about gold that were rife in the New York banks. Appendix 2's table for the event analysis also reports a test designed to see if relaxing the requirement that markets clear “instantaneously,” in favor of an approach that allow them some days to assimilate fully the impact of the news would prove fruitful. It did—as can be seen by the results for the pooled residuals for the rest of the week (reported below the main table for the event analysis in Appendix 2).Google Scholar

¹⁷ Their analysis is also flawed by its reliance on a model of the stock market's valuation of equities derived from Paul Samuelson's old paper on asset duration. Our earlier article's point was that such an approach is insufficient for analyzing banks in a situation like 1932. It is not enough that the weighted sum of future profits add to some positive number—one has to reckon with the possibility that some part of the stream sinks low enough to throw the firm into bankruptcy. As we argued there, the question is indeed an empirical one. Samuelson, however, abstracted from this possibility, while our analysis underscored the role that liquidity, bank runs, and balance sheet compositions play in situations such as that of 1932. See Samuelson, Paul, “The Effect of Interest Rate Increases on the Banking System,” American Economic Review, 35 (03. 1945), pp. 16–27.Google Scholar The original work on asset durat on was done by McCauley, F. R. in the 1930s; our discussion of his and other contributions had to be deleted from the published version of our article for reasons of space.Google Scholar

¹⁸ On the role of exchange-rate regimes, see, for example, Temin, Peter and Wigmore, Barrie, “The End of One Big Deflation,” Explorations In Economic History, 25 (10 1990).Google Scholar

¹⁹ Baruch, Bernard to Owen D. Young, April. 16, 1932 (Baruch papers, Seeley Mudd Library, Princeton University). Baruch during this period practiced what he preached: he was buying gold. During the Franklin Roosevelt administration his efforts to hang on to his hoard became a minor scandal.Google Scholar

²⁰ See Ferguson, Thomas, “Normalcy to New Deal: Industrial Structure, Party Competition, and American Public Policy in the Great Depression,” International Organization, 38 (Winter 1984), pp. 79ff.CrossRefGoogle Scholar

²¹ The results of those tests are presented in Appendix 2. Note that under a gold standard, expansionary open market operations can normally be expected to trigger a gold loss; this point is separable from anyone's views about what was desirable economic policy and what the stock market itself was doing.Google Scholar

²² The New York Times, July 29, 1932; the discussion is about the “cumulative evidence of the great change which has occurred in financial sentiment since the atmosphere of gloom seemed all-pervasive two months ago.”Google Scholar

²³ McDougal had served in his post almost two decades and had originally been tapped for it because of his reputation as one of the most able and experienced of Chicago's bankers. In addition, as our article noted, his was not the only voice complaining about the impact of rates on profitability within the banking community. Indeed, a careful study of the Depression will show that issue surfacing again and again. For McDougal, see James, , Chicago Banks, vol. 2, p. 877;Google Scholar for other governors, see the remarks of Governor Norris of Philadelphia, quoted in Epstein and Ferguson, “Monetary Policy,” p. 982. Though it exceeds the scope of this reply, the issue of excess reserves simmered throughout the early New Deal; see, for example, the brief discussion in Ferguson, “Normalcy,” pp. 90–91, of the 1936 election, in which after the Federal Reserve Board raised the reserve requirement, the outgoing president of the American Bankers Association endorsed Roosevelt's re-election bid.Google Scholar

²⁴ Appendix I describes the sources for our calculations of bank margins. Note that the full effect of the plunge in T-bill rates shows up only some months after the initial fall, as older bills paying more run off, and the money has to be reinvested. See Epstein and Ferguson, “Monetary Policy,” pp. 970–72.Google Scholar

²⁵ Estimates of bank profits during the period in question can be found in the Federal Reserve Bulletin, February 1938, pp. 102–26, especially p. 116, table 1.Google Scholar

²⁶ Our quotation follows the Commercial and Financial Chronicle, January 7, 1933, p. 3, which reprinted the Chicago Tribune story of “Saturday last”; we have regularized minor bits of punctuation in accord with more modern usage.Google Scholar