He is grateful to two anonymous referees, an anonymous econometrician, and the coeditors for excellent comments, but responsibility for the contents rests with the author alone. This paper was completed while the author was Visiting Professor of Economics at the Graduate School of Business, University of Chicago.Google Scholar

¹ Two current theories are not tested, the exchange-market-pressure hypothesis and the asset approach; the former because its primary applicability is to a managed rather than clean float, the latter because the requisite data on foreign-asset holdings are lacking for the greenback period.Google Scholar

² Appendices describing data sources, construction of variables, and tests of the quality of these time series are available from the author on request. A copy of these appendices is in the archives of this JOURNAL.Google Scholar

³ Although the cleanness of the float simplifies the testing of exchange-rate theories, the expectation that the dollar would return to a known, fixed exchange rate—the antebellum parity—was a complicating factor.Google Scholar

⁴ In contrast, much work has been done on the Canadian floating dollar from 1950 to 1962, which was not much different from the greenback period and where the data situation is excellent.Google Scholar

⁵ It is interesting that among the many commentators on this episode, only Hammond has noted that the action of the banks was illegal. See Hammond, Bray, Sovereignty and an Empty Purse (Princeton, 1970), p. 157.Google Scholar

⁶ Mitchell, Wesley Clair, A History of the Greenbacks (Chicago, 1903), Pt. 1, chap. 1;Google ScholarBarrett, Don C., The Greenbacks and Resumption of Specie Payments, 1862–1879 (Cambridge, MA, 1931), pp. 3–15;CrossRefGoogle ScholarKindahi, James K., “Economic Factors in Specie Resumption: The United States, 1865–79,” Journal of Political Economy, 69 (02 1961), 31;Google ScholarHammond, Sovereignty, chaps. 1–5;Google ScholarThompson, Gerald Richard, “Expectations and the Greenback Rate, 1862–1878,” Ph.D. dissertation, University of Virginia, 1972, pp. 22–28.Google Scholar

⁷ Friedman, Milton and Schwartz, Anna Jacobson, A Monetary History of the United States 1867–1960 (Princeton, 1963), pp. 59–60, fn. 64, and p. 66.Google Scholar

⁸ A third, unusual, explanation of suspension is provided by Timberlake, Richard H. JrThe Origins of Central Banking in the United States (Cambridge, MA, 1978), pp. 85–86. He argues that government issues of paper money (“U.S. notes”) expanded the money supply and therefore caused prices to rise. The new currency issues provided additional reserves to the banks, which proceeded to create deposits and issue more bank notes, involving still further rises in the price level. Private parties saw the equilibrium gold price as above its mint price. U.S. notes, as well as bank notes and deposits, were redeemed for gold, and suspension of specie payments resulted. Timberlake's explanation is dubious, apparently resting on a confusion of dates. The only government paper currency extant prior to suspension were the Treasury demand notes, whereas his account of events involves the actually post-suspension, larger, issues of U.S. legal-tender currency. Indeed, Timberlake's explanation is illogical for a second reason: Treasury demand notes could not be used as bank reserves.Google Scholar

⁹ Comprehensive discussions of the political and economic aspects of the issue of the greenbacks are provided by Mitchell, A History, Pt. 1, chaps. 2–5; Barrett, The Greenbacks, PP. 15–73;Google ScholarHammond, Bray, “The North's Empty Purse, 1861–1862,” American Historical Review, 67 (10 1961), 1–18; andCrossRefGoogle ScholarHammond, Sovereignty, chaps. 6–12.Google Scholar

¹⁰ Timberlake errs in stating that greenbacks returned to the Treasury in payment for tariffs. See The Origins of Central Banking, p. 86.Google Scholar

¹¹ The effect of specie suspension and greenback issue on the circulation of the treasury demand notes is described by Mitchell, A History, pp. 149–56, 180, and Barrett, The Greenbacks, pp. 74–75.Google Scholar

¹² See Friedman and Schwartz, A Monetary History, pp. 16–23, andGoogle ScholarWest, Robert Craig, Banking and Reform and the Federal Reserve, 1863–1923 (Ithaca, 1977), chap. 1, for discussions of the formative years of the national banking system.Google Scholar

¹³ See Mitchell, A History, pp. 174–78, and Barrett, The Greenbacks, pp. 77–78, for histories of the interest-bearing currencies of the Civil War.Google Scholar

¹⁴ See Mitchell, A History, p. 142; Barrett, The Greenbacks, p. 73; andGoogle ScholarLester, Richard A., Monetary Experiments (Princeton, 1939), pp. 161–62. Only Lester refers to Gresham's law by name.Google Scholar A later writer with this view is Timberlake, Richard H. Jr “The Resumption Act and the Money Supply,” Journal of Monetary Economics, 1 (07 1975), 343.CrossRefGoogle Scholar

¹⁵ After various makeshift substitutes for subsidiary silver coins, the treasury issued paper fractional currency (essentially greenbacks in denominations under $1) under the third legal-tender act. For accounts of this small-change episode and related ones involving one-to-five-cent coins, see Mitchell, A History, pp. 156–73; Barrett, The Greenbacks, pp. 75–77; and Lester, Monetary Experiments, p. 162.Google Scholar

¹⁶ This difference between the effect of the greenback on the circulation of gold and silver coin was pointed out by Friedman and Schwartz, A Monetary History, p. 27.Google Scholar

¹⁷ For an institutional description of the gold market that developed after suspension, see Mitchell, A History, pp. 182–87;Google ScholarMitchell, Wesley Clair, Gold, Prices and Wages under the Greenback Standard (Berkeley, 1908), pp. 1–3; andGoogle ScholarThompson, Expectations, pp. 32–35. In June 1864 Congress legislated the prohibition of the gold market (though allowing brokers to transact in gold within their offices), but the bill was repealed in July. The intent was to reduce the premium on gold, but the law was ineffective, with the premium actually increasing.Google Scholar

¹⁸ See Mitchell, A History, p. 182.Google Scholar

¹⁹ Mitchell, A History, pp. 142, 180; Barrett, The Greenbacks, pp. 55, 73; and Lester, Monetary Experiments, p. 162, all err in asserting that no gold corn remained as circulating medium in the North, although Mitchell and Barrett qualify their statements. Friedman and Schwartz, A Monetary History, p. 17, provide detailed data on components of the money supply in mid-1867. They show combined gold-coin holdings of $48 million for commercial banks and the public.Google Scholar

²⁰ According to Friedman and Schwartz, A Monetary History, p. 17, a total of $19 million of gold certificates were held by the banks and the public in mid-1867.Google Scholar

²¹ See Friedman and Schwartz, A Monetary History, pp. 26, 28–29, fn. 17, and Kindahl, “Economic Factors,” 32, fn. 8. The law permitted either greenback dollars or gold coin to serve as bank reserve assets against deposits, but established a dollar-for-dollar fixed exchange rate between them for this purpose alone. So a kind of Gresham's Law operated in the case of bank reserves. Sound banking practice mandated a sufficient gold reserve to cover gold deposits, but not much more than that, in view of the higher value of gold relative to greenbacks on the free market.Google Scholar

²² Lester, Monetary Experiments, pp. 163–66, in the classic treatment of the retention of the gold standard in California and Oregon, explains why the greenback did not become the primary money in the West.Google Scholar

²³ Mitchell, Gold, Prices and Wages, pp. 251–58.Google Scholar

²⁴ Kindahl, “Economic Factors,” 33, fn. 12, comments that even after the Civil War, when the gold premium became more stable, there is no evidence of a return to the former practice of quoting foreign exchange directly in greenbacks.Google Scholar

²⁵ For example, See Taussig, F. W., “International Trade Under Depreciated Paper: A Contribution to Theory,” Quarterly Journal of Economics, 31 (05 1917), 384–85;CrossRefGoogle ScholarKindahl, “Economic Factors,” 32–33; Friedman and Schwartz, A Monetary History, pp. 58–61; Thompson, Expectations, pp. 4–7; andGoogle ScholarWimmer, Larry T., “The Gold Crisis of 1869: Stabilizing or Destabilizing Speculation under Floating Exchange Rates?” Explorations in Economic History, 12 (04 1975), 106–07.CrossRefGoogle Scholar

²⁶ See Davis, L. E. and Hughes, J. K. T., “A Dollar-Sterling Exchange, 1803–1895,” Economic History Review, 13 (08 1960), 52–78, and, in support of their position,Google ScholarYeager, Leland B., International Monetary Relations: Theory, History, and Policy (New York, 1976), p. 297.Google Scholar

²⁷ See Barrett, The Greenbacks, chaps. 5–10; Kindahl, “Economic Factors”; Friedman and Schwartz, A Monetary History, pp. 29–50, 79–85;Google ScholarUnger, Irwin, The Greenback Era (Princeton, 1964);CrossRefGoogle ScholarTimberlake, Richard H. Jr “Ideological Factors in Specie Resumption and Treasury Policy,” this JOURNAL, 24 (03 1964), 29–52;Google ScholarTimberlake, “The Resumption Act”; and Timberlake, The Origins, chaps. 7–8.Google Scholar

²⁸ Mitchell, A History, pp. 198–238; Mitchell, Gold, Prices and Wages, pp. 14–15; Thompson, Expectations, chaps. 3–5.Google Scholar

²⁹ See Fisher, lrving, The Theory of Interest (New York, 1930), pp. 401–02;Google ScholarFriedman and Schwartz, A Monetary History, pp. 69–76; andGoogle ScholarRoll, Richard, “Interest Rates and Price Expectations During the Civil War,” this JOURNAL, 32 (06 1972), 476–98.Google ScholarThis position is also argued by Fels, Rendigs, American Business Cycles 1865–1897 (Chapel Hill, 1959), p. 59.Google Scholar

³⁰ See Friedman and Schwartz, A Monetary History, pp. 69–76, and Wimmer, “The Gold Crisis.” The latter author, in a nontraditional account of Jay Gould's manipulation of the gold market in August-September 1869, argues that speculative capital flows were the primary force maintaining the stability of the foreign-exchange market and therefore countering Gould's attempt to increase the greenback price of gold.Google Scholar

³¹ It can be stated legitimately that even a year may be insufficient time for the full response of trade to exchange-rate changes to work itself out. But if lagged relationships are to be investigated, a monthly or quarterly observation unit should be used, and intraannual data are not available for several specified variables in the model.Google Scholar

³² Friedman and Schwartz, A Monetary History, p. 15. In fact, there are only two known cases of government intervention. On September 24, 1869 (so-called Black Friday), the Treasury sold 4 million gold dollars in a successful effort to break Jay Gould's upward manipulation of the gold premium. Beginning March 1877, the Treasury pursued policies to increase its gold reserve in order to enhance prospects for a successful resumption of specie payments.Google Scholar

³³ In 1862–1879 (years ending June 30), 52 percent of U.S. merchandise exports went to Britain and 35 percent of U.S. imports came from Britain. No other country was the destination of as much as 10 percent of U.S. exports and none (with the exception of Cuba) was the source of as much as 10 percent of U.S. imports. The data source is U.S. Bureau of the Census, Historical Statistics of the United States (Washington, D.C., 1975), 904, 906–07.Google Scholar

³⁴ The PPP theory augmented with capital flow could not be formulated in logarithmic form, because K can be negative.Google Scholar

³⁵ The method of estimation used is the Cochrane-Orcutt iterative technique, which corrects for autocorrelation. To the extent that such autocorrelation is due to the presence of speculative influences and dynamic adjustment factors that remain even with annual observations, the technique corrects for these elements.Google Scholar

³⁶ The data appendices are available from the author on request.Google Scholar

³⁷ The year 1860 satisfies the criterion of a “normal” base period; it is the last full year that predates the Civil War. Also, a good case can be made that 1860 was a year of balance-of-payments and exchange-market equilibrium: the deviation of the market gold-dollar/sterling rate from the mint parity was the lowest it would be until 1877. Indeed, this deviation climbs from less than 1/10th of one percent in 1860 to almost 1–3/4 percent in 1861.Google Scholar

³⁸ The linear Keynesian equation is excluded because inclusion of the income variable lowers the equation's explanatory power.Google Scholar

³⁹ The exceptions are not really damaging to the PPP theory because with a second explanatory variable, one should allow for the possibility that the constant term might be affected.Google Scholar

⁴⁰ Considering the two linear augmented-PPP equations, the reason why the interest-rate effect is so much more significant than, and also almost double the magnitude of, the capital-flow effect might be the poor quality of the available data on international capital movements.Google Scholar

⁴¹ Taussig, “International Trade”; Graham, F.D., “International Trade under Depreciated Paper: The United States, 1862–79,” Quarterly Journal of Economics, 36 (02 1922), 220–73;CrossRefGoogle ScholarKindahl, “Economic Factors”; Friedman and Schwartz, A Monetary History;Google ScholarFarag, Attiat A. and Ott, David J., “Exchange Rate Determination under Fluctuating Exchange Rates: Some Empirical Evidence,” in Murphy, J. Carter, ed., Money in the International Order (Dallas, 1964), pp. 84–105;Google ScholarThompson, Expectations;Google ScholarAldrich, Mark, “Flexible Exchange Rates, Northern Expansion, and the Market for Southern Cotton: 1866–1879,” this JOURNAL, 33 (06 1973), 401–05.Google Scholar

These studies all adopt an augmented PPP approach in one form or another, but suffer from use of (i) the wholesale price index rather than the broader-based GNP deflator as the price concept for PPP and (ii) the greenback price of gold rather than the pound/greenback rate to represent the exchange rate. Further, Farag and Ott are the only authors to fit PPP regressions involving interest rates; their results are poor, probably because they intermix the short and long run by using semiannual observations and including both short-term and long-term interest-rate variables.Google Scholar

⁴² See Friedman and Schwartz, A Monetary History, p. 63, fn. 67.Google Scholar

⁴³ Using capital-flow data, Kindahl further adjusts the limits, a procedure ignored here because of its subjective nature. See Kindahl, “Economic Factors,” p. 37.Google Scholar

⁴⁴ Also, of course, there must have been no structural changes in the U.S. and British economies since the base period. Aldrich, “Flexible Exchange Rates,” modifies the Kindahl analysis by incorporating structural changes in international trade. After the Civil war, such changes favored U.S. comparative advantage in new goods and helped explain the higher real exchange value of the greenback. This point is also made by Friedman and Schwartz, A Monetary History, p. 78.Google Scholar

⁴⁵ A Monetary History, pp. 58–78.Google Scholar