Chapter 1: Complications for the United States from International Credits: 1913–401

Thomas Sargent, George Hall, Martin Ellison, Andrew Scott, Harold James, Era Dabla-Norris, Mark De Broeck, Nicolas End, Marina Marinkov, and Vitor Gaspar
Published Date:
November 2019
  • ShareShare
Show Summary Details
George J. Hall and Thomas J. Sargent 

The granting of foreign credit is a first step toward the establishment of an aggressive foreign policy.

Adams (1887, 25)

Money is the worst of all contrabands. … I know of nothing that would do more to prevent war than an international agreement that neutral nations would not loan to belligerents.

Secretary of State William Jennings Bryan to President Woodrow Wilson,

August 10, 1914

We are going into war upon the command of gold.

Senator George Norris (speech before the Senate, April 4, 1917, 65th Congress, 1st Session)

It is the instigators of this war who deserve to bear this lead weight of billions. Let them drag it through the decades to come, not us.

Dr. Karl Helfferich, Secretary of State for the Treasury (August 20, 1915, speech before the Reichstag)

And forgive us our debts, as we forgive our debtors.

Matthew 6:12, King James Bible

It is highly improbable that Congress or popular opinion in this country will ever permit cancellation of any part of the debt of the British Government to the United States as an inducement towards a practical settlement of the reparation claims.

Woodrow Wilson to David Lloyd George, November 2, 1920


Before 1914, the US was a net debtor to foreigners. But in 1887, the American economist Henry Carter Adams had forecast that US economic growth and falling returns on US securities would eventually transform the US into a net foreign creditor. Adams said that would threaten George Washington’s “true policy to steer clear of permanent alliance with any portion of the foreign world.”

It lies altogether within the range of possibilities that the city of New York, like the cities of London and Paris, should become a storehouse of capital to which the sovereigns of petty states may resort to fill their depleted treasuries. This tendency is fraught with danger to the policy of isolation thus far maintained by the United States, and it becomes an important question, what attitude this country should assume with regard to the interests of those who place their funds beyond the control of American law. One of two policies must be declared, nor ought the nation to be permitted to drift in this matter. Either citizens of this Republic should know that money placed in foreign bonds is at their own risk, or they should prepare themselves to see questions of foreign policy become much more important than they now are. It seems, then, from whichever point of view we consider the question, that the United States can not reasonably expect to avoid political complications sure to come with an extension of international credits; and it is on this account desirable that the Federal Government should present a clearly formulated policy, upon which the public may rely.

Adams (1887, 37–38)

We do not know whether Professor Woodrow Wilson had read Henry Carter Adams’s 1887 book, but we do know that Secretary of State William Jennings Bryan conveyed the essence of Adams’s message to President Wilson in an August 10, 1914, letter that we reproduce in Anne 1.2. President Wilson rejected Secretary Bryan’s advice to prohibit US citizens from lending to belligerents. Instead, in 1914 Wilson urged but did not compel US citizens to be “impartial in thought as well as in action.” American banks and other investors ignored Wilson’s advice and bought billions of dollars of UK and French government bonds.2 J.P. Morgan & Co. marketed those bonds to US citizens and also served as sole agent for the British and French governments when they purchased billions of dollars of war supplies from US producers.

When war began in August 1914, the UK was a net creditor to the US. During the war, the UK government forced its citizens to exchange their US securities for British sovereign debt and then used those securities as collateral for sovereign UK bonds sold to American citizens. By the late fall of 1916, the UK government had nearly exhausted that collateral. Therefore, on November 27, 1916, at the urging of President Wilson, who by then had reconsidered his rejection of Secretary Bryan’s August 1914 advice, the Federal Reserve Board publicly warned US citizens not to buy more British or French government debt.3 UK citizens and civil servants including John Maynard Keynes welcomed the Federal Reserve Board’s message because they were then using the UK government’s financial distress to strengthen their recommendation that the UK accept what Woodrow Wilson would soon call a “peace without victory.”4

But hawkish British and German government officials interpreted the November 27 Federal Reserve memorandum as only a temporary setback to British credit in America and focused instead on what they recognized as the same decisive interest that Henry Carter Adams had identified in 1887: private US creditors of the UK and French governments and US export producers who had benefited from those credits wanted Entente victory. Opponents of peace without victory in both the UK and Germany pointed to those American interests. Hawkish Germans argued that, because American private creditors had lent so much to Entente governments, the US would soon enter the war on the Entente side, whether or not Germany began unrestricted submarine warfare.5 Germany resumed unrestricted submarine war on February 1, 1917, and the US entered the war on April 6, 1917. The US thus failed to “avoid political complications sure to come with an extension of international credits” as forecast by Henry Carter Adams in 1887.6

Complications from International Credits

Before the US entered the war on April 6, 1917, it was private US bondholders who had lent to the British and French governments. Afterward, it was the US government. On April 24, 1917, President Wilson signed the First Liberty Loan Act. It authorized the Treasury to borrow up to $5 billion and to purchase up to $3 billion of Entente and Allied debts. Britain and France soon refinanced their short-term debts to private US citizens by borrowing from the US government, which in turn borrowed from US citizens.7 Congress eventually passed three more Liberty Loan Acts and one Victory Loan Act. Table 1.1 describes features of the Liberty Loan bonds.8 By making them convertible, Congress insured purchasers of First and Second Liberty Loan bonds against risk that interest rates would rise. Purchasers of these bonds could convert them at par into new bonds with the same maturity date and call provisions as their original loan but with the coupon rate and tax provisions of subsequent issues. Higher later coupon rates did indeed induce many owners of First and Second Liberty Loan bonds to exercise those conversion options. Of the nearly $2 billion First Liberty Bonds sold, about $560 million were converted into higher coupon-paying bonds. Of the nearly $4 billion Second Liberty Bonds sold, nearly all were ultimately converted.9

Table 1.1.Liberty and Victory Loans
Coupon RateIssue DateCall DateMaturity DateConvertible?Issued (billions)
First Liberty LoanJun 1917Jun 1932Jun 1947Yes 1$2.0
Second Liberty Loan4Nov 1917Nov 1927Nov 1942Yes 13.8
Third Liberty LoanMay 1918Sep 1928No4.2
Fourth Liberty LoanOct 1918Oct 1933Oct 1938No7.0
Victory Liberty Loan23¾, 4¾May 1919Jun 1922May 1923No4.5
Source: US Treasury Monthly Statement of the Public Debt.

If the Treasury issued a new series of bonds at a higher interest rate prior to the end of the war, investors in the First and Second Liberty Loans could exchange their bonds at par for new bonds with the maturity date and call provisions of their original series but with the coupon rate and tax exemptions of the subsequent series. There was no limit to the number of times holders of the First Liberty Loan could exercise this option, but holders of the Second Liberty Loan could exercise this option only once.

The Victory Liberty Loan paid a 3¾ coupon per annum if the coupon payments were exempt from federal income taxes; otherwise, this loan paid a 4¾ coupon per annum.

Source: US Treasury Monthly Statement of the Public Debt.

If the Treasury issued a new series of bonds at a higher interest rate prior to the end of the war, investors in the First and Second Liberty Loans could exchange their bonds at par for new bonds with the maturity date and call provisions of their original series but with the coupon rate and tax exemptions of the subsequent series. There was no limit to the number of times holders of the First Liberty Loan could exercise this option, but holders of the Second Liberty Loan could exercise this option only once.

The Victory Liberty Loan paid a 3¾ coupon per annum if the coupon payments were exempt from federal income taxes; otherwise, this loan paid a 4¾ coupon per annum.

Between 1917 and 1920, the US War and Navy departments spent $20 billion and the Treasury extended $9.5 billion in loans to Britain, France, Italy, and eight other allies of Entente countries.10 After that, the Treasury extended credits to an even larger group of countries in order to finance US sales of surplus war materials and US relief supplies to Europe.11 By December 1922, a total of 20 nations owed the Treasury $11.8 billion ($10.1 billion in principal and $1.7 billion in interest in arrears). The face value of those foreign loans represented 52 percent of privately held US federal debt and 16 percent of US GDP in 1922.12

Table 1.2 summarizes a network of international debts and credits that had emerged at the end of the war and that framed political questions that would preoccupy statesmen and citizens for the next 15 years. Should debts be paid? By whom? Should price levels and exchange rates of currencies in which loans were denominated be adjusted to redistribute resources between creditors and debtors within and across countries? Should governments discriminate between domestic and foreign creditors?13

Table 1.2.Interallied Indebtedness at the Armistice (in millions of dollars)
DebtorTo USTo BritainTo FranceTotal
Great Britain3,6963,696
Jugoslavia (Serbia)1192297400
Source: War Debt Supplement, The Economist, November 12, 1932, page 2.
Source: War Debt Supplement, The Economist, November 12, 1932, page 2.

During the war, both German and French finance ministers had answered the who-should-pay question. The German finance minister assured the Reichstag that the “lead weight of billions” would be carried by Germany’s enemies.14 Meanwhile, the French finance minister assured the French parliament that France’s enemies, not Germany’s, would service French war bonds.

With more foresight, a wartime French finance minister could have said that the burden of paying French war bonds would fall either on France’s enemies or on its foreign friends or maybe on its domestic creditors and taxpayers. Deciding would preoccupy Europeans and Americans from the time of Armistice on November 11, 1918, until the trough of the world depression in 1933. Failed improvisations postponed a permanent settlement until 1933. In 1919, John Maynard Keynes had urged immediate agreement to what would ultimately be the disposition of postwar international loans and reparations.15 Keynes said that the US should write down its loans to the UK, France, and Italy in exchange for their accepting smaller reparations from Germany. Keynes said that would reset national balance sheets in ways that would foster reconstruction of international monetary and trading systems. It is true that in 1922 Congress established the World War Foreign Debt Commission to write down and reschedule outstanding Treasury credits to foreign governments, but not until 1933 would governments accept the comprehensive adjustments that Keynes had recommended in 1919. By then a fragile monetary and trading system assembled in the 1920s had collapsed because it had depended on what turned out to be too optimistic assumptions about both macroeconomic growth and politically sustainable fiscal-monetary policies.

Purpose and Organization

This chapter constructs a quantitative account of Henry Carter Adams’s “political complications sure to come with an extension of international credits” by the US during World War I, the absence in 1914 of “a clearly [US] formulated policy, upon which the public may rely,” outcomes brought by the absence of such a policy, and their consequences for US federal monetary and fiscal arrangements in the 1920s and 1930s.16 We apply an accounting framework described in Anne 1.1.17 We work in the Davis Dewey (1912) tradition of spotlighting objects in government budget constraints and balance sheets. These objects frame financial questions created by Adams’s “political complications” and our answers to them: How large were the US government credits extended to foreign governments during the war? How big were the promised and realized federal income streams generated by these credits, and how did those quantities compare with interest payments on federal debts issued during the war? How did defaults, reschedulings, and repudiations of foreign credits and fluctuations in rates of growth of GDP, nominal interest rates, inflation, and primary federal government budget surpluses contribute to the evolution of the federal debt-to-GDP ratio?18

Although we focus mainly on the period during and after World War I, we also compare outcomes with those from a period of similar length: during and after the US Civil War. The Civil War experience set precedents that US policymakers naturally consulted during and after World War I.19,20 Two 1913 reconstructions of US fiscal-monetary infrastructure distinguished the situations confronting Congress during the Civil War and World War I: the Federal Reserve System and the 16th Amendment to the US Constitution and the Revenue Act of 1913 that authorized a federal income tax. The US used both of these new institutions to help finance World War I. By way of contrast, the US entered the Civil War without a federal income tax and without either a central bank or a national banking system. Although Civil War Congresses levied a federal income tax and constructed a new National Monetary System, it took time to get the administrative machinery up and running. The US entered World War I with a Federal Reserve System and a federal income tax ready to go, having only to choose monetary policy actions and set income tax rates within established institutions.21

Debts, Credits, Values, and Returns


Figures 1.1, 1.2, and 1.3 set the scene.22Figure 1.1 shows the big rise in military expenditures as well the granting of large foreign credits during the war. The war brought a permanent rise in federal expenditures as a fraction of GDP. Figure 1.2 shows that while federal revenues as a fraction of GDP rose during the war, they rose much less than expenditures. That discrepancy gave rise to the wartime growth of federal government debt depicted in Figure 1.3.23 The book values in Figure 1.3 show that during the war the US Treasury borrowed large amounts and lent large sums to European combatants.24 Book values indicate that net US government indebtedness was actually substantially less than the gross amount depicted in Figure 1.3 because large fractions of outstanding US Treasury bonds were “backed” by claims on foreign governments. How much “backing” those credits would in fact put behind US Treasury bonds, and how much relief they would bring the US taxpayers who were ultimately responsible for servicing those bonds, would depend on how faithfully those foreign debtors honored their obligations to the US Treasury. In the end, they would pay only small fractions of what they had promised, but it would take a long time for all parties to accept that.

Figure 1.1.US Federal Expenditures by Type (Percentage of GDP)

Source: Foreign credits, United States World War Foreign Debt Commission (1927), Exhibit 116, pp. 318–325; other expenditure series, Annual Report of the Secretary of the Treasury on the State of the Finances, 1941, pp. 408–416; GDP,

Figure 1.2.US Federal Revenues by Types (Percentage of GDP)

Source: Revenue series, Annual Report of the Secretary of the Treasury on the State of the Finances, 1941, pp. 408–416; GDP,

Figure 1.3Face Value of US Treasury Debt and Face Value of Foreign Securities Held by the Treasury (Including Accrued Interest)

Source: US Treasury Monthly Statement of the Public Debt, and Board of Governors of the Federal Reserve System (1943), Table 149.

Note: US Treasury debt excludes securities held in US government accounts and by the Federal Reserve.

We want to know how accurately the book values in Figure 1.3 approximate market values.25Figure 1.4 shows that book and market values of outstanding Treasury bonds approximate each other pretty well. This need not have occurred; it is a consequence of the way the Treasury managed its debt in light of market interest rates. In a later section of this chapter, titled “Bond Prices and Quantities,” we describe how we use prices and quantities of individual bonds to measure the total market value of Treasury debt.

Figure 1.4.Ratio of Market Value to Par Value of Privately Held US Treasury Debt

Source: Hall et al. (2018) and author calculations.


Figures 1.5 and 1.6 show two ways of depicting nominal and real returns on the Treasury bond portfolio. Figure 1.5 shows annual nominal and real net returns, whereas Figure 1.6 shows the real and nominal outcomes of purchasing $100 of the bond portfolio in January 1917 and continually reinvesting proceeds in a rebalanced portfolio of Treasury securities. Figure 1.6 reveals how US price level movements influenced real returns.

Figure 1.5.Nominal and Real Returns on the US Treasury’s Bond Portfolio

Source: Hall et al. (2018) and author calculations.

Figure 1.6.Nominal and Real Values of $100 Invested in January 1917 in US Treasury’s Bond Portfolio

Source: Hall et al. (2018) and author calculations.

Price Level Movements

In Figure 1.3, both federal debts and foreign credits are denominated in US dollars, so the US price level shaped the value of those debts and credits in terms of goods and services. Figure 1.7 indicates that the logarithm of the US price level rose especially rapidly after the US entered the war in 1917 and then shows a sharp drop in 1920–21 and a larger and longer one from 1929 to 1933 that nevertheless left the price level in 1933 23 percent above its prewar level. These price level movements coincided with European combatants leaving the gold standard during the war, their postwar implementations of monetary experiments that replaced precious metal monies with paper monies,26 and subsequent modifications and terminations of those experiments carried out during the depression of the 1930s.27 The US, which imperfectly but nominally remained on the gold standard during the war,28 shared a 1920–21 price level drop with European countries that were then deciding at what rates to exchange their currencies for gold, politically charged decisions that would redistribute wealth among domestic private creditors and their domestic debtors. UK domestic creditors who owned pound-denominated claims on the UK government earned high real returns when the UK restored convertibility of the pound to gold at the prewar par of 4.80 dollars per pound in April 1925, whereas domestic creditors of the French government who owned franc-denominated claims fared badly when France ultimately resumed convertibility of the franc to gold at 20 percent of its prewar value. The period of slowly rising US prices from 1922 to 1929 in Figure 1.7 coincided with the UK’s success in temporarily establishing an international “gold exchange standard” that decreased monetary demands for gold by making IOUs denominated in British pounds a reserve currency for British Empire Dominions and colonies and much of Europe as well.

Figure 1.7.Natural Log of US Price Level

Source: National Bureau of Economic Research, Indicator m04051.

Nominal versus Real Returns

While reschedules, defaults, and repudiations shaped dollar values of the US Treasury’s credits, US inflation shaped real values of those credits and also of the US Treasury’s debts. By affecting both the price level movements in Figure 1.7 and the US government debt and credit dynamics in Figure 1.3, US and foreign monetary-fiscal policies influenced the rate of return outcomes summarized in Figures 1.5 and 1.6. Figure 1.5 shows nominal and real one-period returns on a value-weighted rebalanced portfolio of US Treasury bonds, whereas Figure 1.6 shows cumulative nominal and real values coming from continually reinvesting in that portfolio, starting with an initial investment of $100 on January 1, 1917.29Figure 1.6 reveals that the 1920–21 price level drop brought the real value back to its initial value after wartime inflation had reduced it by one-fourth. Creditors of the US government earned high real returns from 1929 to 1933 and low ones from 1933 to 1937.

Comparison with Civil War

During the 1920s and 1930s in both the US and in Europe, the distinction between nominal and real returns on government bonds was on the minds of policymakers and bondholders. For example, Treasury Secretary Andrew Mellon wrote:

The real value of the dollar, that is, its value in terms of goods it will purchase, does not remain constant. The experience with our Civil War debt was that we borrowed a 54-cent dollar and repaid an 85-cent dollar (using the 1860 value as the base) or, in other words, we paid back in value $3 for every $2 we borrowed. Using 1913 as a base, our present war debt was borrowed on a 51-cent dollar, and to-day the dollar is worth 66 cents. If the appreciation of the dollar continues—and such has been fiscal history after other great wars—then the longer we postpone payment the more in real value we will have to pay.

Mellon (1926, 7)

Like Secretary McAdoo before him,30 Secretary Mellon sought lessons from the Union’s fiscal and monetary policies implemented during the US Civil War.

Figure 1.8 compares the logarithm of the US price level during and after World War I with a period of the same length during and after the US Civil War. The US price level had returned nearly to its prewar level 20 years after the start of the Civil War in 1861; 20 years after the US had entered World War I in 1917, the price level was still 63 percent above its 1914 level. But in 1933, the same gold standard discipline that had pushed the price level to its prewar value after the Civil War seemed to be driving it back toward its pre–World War I level.31 The Roosevelt administration did not want the wealth redistributions that this would have brought, so it pursued a policy designed to restore the price level to its 1928 or 1929 level.32

Figure 1.8.Natural Log of US Price Level during and after Two Wars

Source: National Bureau of Economic Research, Indicator m04051

Note: The ticks on the x-axis correspond to August for the 1856–84 period and January for the 1912–40 period.

There were important differences in international and domestic monetary arrangements during the two postwar episodes compared in Figure 1.8. During the US Civil War, the UK remained on the gold standard while the United States left it and made a paper currency called the greenback a legal tender and unit of account in which the log price levels in Figure 1.8 are recorded. After 1862 and before the US Treasury returned to gold in 1879 by promising to trade one greenback dollar for one gold dollar, one gold dollar traded for more than one greenback dollar. By way of contrast, during and after World War I, it was the UK and other European countries that had abandoned the gold standard, while the US, until 1933, more nearly remained on gold.33 Europe’s leaving gold lowered the relative price of gold by decreasing world demand for gold for monetary uses; that changed the price of gold relative to most goods and services, and that in turn exported inflation from Europe to the US.34 The UK temporarily restored the pound to its prewar exchange rate in 1925, but then left the gold standard permanently in September 1931. By December 1931, one UK pound had fallen from its value of $4.87 in August 1931 to about $3.7.

Figure 1.9 compares real rates of return on the US government bond portfolio across the post–Civil War and the post–World War I periods.35 US Treasury creditors suffered similar cumulative real losses during both wars, and although they enjoyed cumulative gains after both wars, they fared better 20 years after the Civil War than they did 20 years after World War I. Comparing the log price levels in Figure 1.8 with the cumulative real returns in Figure 1.9 indicates how price level movements helped award bondholders bigger cumulative real returns after the Civil War than after World War I. Not wanting to let the price level sink to its pre–World War I level, the Roosevelt administration promoted a monetary policy that aimed to restore the price level to its 1929 level.36 Because he rejected international restrictions on US monetary policy, in late June 1933, President Roosevelt directed the US delegation to the London Economic Conference not to assist international efforts temporarily to stabilize exchange rates of the US dollar to the currencies of France, the UK, and other countries. Anne 1.4 reproduces Roosevelt’s “bombshell message” to the conference in which he rejected superficial temporary measures in favor of ones designed permanently to restore the economic fundamentals on which a gold standard rests.37 For President Roosevelt, those fundamentals included balanced national government budgets.

Figure 1.9.Real Values of $100 Invested in January 1861 and January 1917 in US Treasury’s Bond Portfolio

Source: Hall et al. (2018) and author calculations.

Figure 1.10.Real Returns on Treasury Portfolio during and after Two Wars, Annual by Fiscal Year

Source: Hall et al. (2018) and author calculations.

Different Price-Output Outcomes after Two Wars

Figures 1.11 and 1.12 add measures of real output to the price levels presented in Figure 1.8. There is a strong positive association between the price level and industrial output in the post–World War I period in Figure 1.12, but not in the post–Civil War period depicted in Figure 1.11. In both periods, contemporary commentators and statesmen complained about adverse deflation-induced redistributions from debtors to creditors that are the heart of Irving Fisher’s (1933) debt-deflation theory of macroeconomic contractions.38 What accounts for the different output-price associations during the two periods? One possibility is that the prevalence of gold clauses on railroad and other corporate bonds after the Civil War meant that bond payouts were indexed against deflation of the greenback, making deflation-induced redistributions much smaller after the Civil War.39 In any event, the pronounced output declines that coincided with aggregate price level declines in the early 1930s convinced President Roosevelt to implement the advice of Irving Fisher and other advocates of price level targeting.40

Figure 1.11.Natural Log of US Price Level and Per Capita Industrial Production (IP) from 1860 to 1879

Source: Price level, National Bureau of Economic Research, Indicator m04051; I P, Davis (2004).

Figure 1.12.Natural Log of US Price Level and Per Capita Industrial Production (IP) from 1910 to 1939

Source: Price level, National Bureau of Economic Research, Indicator m04051; I P, Miron and Romer (1990) and Federal Reserve.

Returns and Government Debt Levels

Figure 1.10 shows annual real returns on the Treasury bond portfolio during and after the Civil War and World War I. Although it conveys less information than Figure 1.9, approximating as it does the time derivative of that figure, Figure 1.10 helps to identify years after each war in which a declining price level boosted real returns, and years after World War I in which a rising price level depressed real returns.41

The immense growth of US government debt during World War I shown in Figure 1.1 is an arithmetic consequence of the discrepancy between the federal government expenditure path shown in Figure 1.1 and the total federal government revenue path shown in Figure 1.2. US World War I finance bears telltale signs of a Barro-Gallatin recommendation to finance net-of-interest deficits during a war with increases in government debt and then, after a war, to run net-of-interest surpluses just sufficient to service the war-enlarged government debt.42 In this respect, the deficit during the depression in the 1930s exhibits fiscal features of a war.43Figure 1.1 shows what appears to be a permanent increase in the level of noninterest government expenditures after the war, even before the onset of the depression in 1930. A permanent increase in federal government expenditures also occurred during the Civil War.

Financial Repression and Subsidies

During and after the Civil War, the Union forced national banks to buy Union bonds (see Dewey 1912). During World War I, the US used a less direct approach. Although it extended short-term credit to the Treasury, the Federal Reserve Banks did not buy bonds directly from the Treasury. Instead, they lent large sums to banks belonging to the Federal Reserve System by discounting loans secured with Liberty Loans as collateral. In this way, the Federal Reserve engineered substantial increases in the monetary base that helped finance the war. Figures 1.13 and 1.14 show the effects of these and other operations during the war, as well as an unwinding of the Federal Reserve’s balance sheet after the war, especially during the 1920–21 recession and the associated worldwide downward movements in price levels that occurred then.44

Figure 1.13.Federal Reserve Assets

Figure 1.14.Federal Reserve Liabilities

The Federal Reserve supported the Treasury in several ways. It directly purchased certificates of indebtedness from the Treasury. Governors of Reserve Banks organized and led committees in each district to sell Treasury bonds. The Federal Reserve supported a program of “borrow and buy” that encouraged individual investors to finance purchases of Liberty Loans by borrowing from their local banks, which could then discount those loans at the Federal Reserve’s discount window. The Treasury designated the New York Federal Reserve as its agent for bond sales. The Federal Reserve lent at preferred discount rates to banks that purchased Treasury certificates of indebtedness. The Federal Reserve lent to member banks at a preferred discount rate if the proceeds were used to purchase Liberty and Victory bonds that the Federal Reserve accepted as collateral.

The Federal Reserve’s balance sheet summarized in Figures 1.13 and 1.14 shows that, through these operations, the Federal Reserve temporarily monetized over a billion dollars of Treasury securities (note the wartime increase in both bills discounted in Figure 1.13 and the increase in Federal Reserve notes in Figure 1.14). This means that during the war, a substantial amount of the Treasury debt in Figure 1.3 was not held by the public but by the Federal Reserve. Figure 1.15 shows time series of pertinent money market rates that determined spreads on these portfolio operations. During the subscription period for the First Liberty Loan from May 15 to June 15, 1917, member banks could borrow from the New York Federal Reserve at 3 percent to buy bonds at par that paid 3.5 percent coupons. As the four Liberty Loan bond drives and the Victory Loan bond drive progressed and as coupon rates on Treasury bonds increased, so did the New York Federal Reserve’s preferred discount rate, but it always remained below the coupon rates and market yields of newly issued Treasury securities. Those spreads motivated US citizens to finance purchases of Treasury bonds by borrowing from banks.

Figure 1.15.New York Federal Reserve Discount Rates Secured by Liberty and Victory Loans, Yields on Liberty and Victory Loans, and Coupon Rates for Liberty and Victory Loans

Friedman and Schwartz (1963) describe how Treasury officials persuaded the Federal Reserve to keep interest rates low in order to help the Treasury sell Victory Loan bonds and how that led the Federal Reserve belatedly to administer excessive interest rate increases that they say worsened the sharp 1920–21 downturn in real economic activity in the US. Although Figure 1.15 shows how the Federal Reserve increased rates during this period, our attention is also drawn to the substantial unwinding of the Federal Reserve’s indirect holdings of Treasury securities reflected in two salient quantities of Figures 1.13 and 1.14, namely, bills discounted and Federal Reserve notes, respectfully. The substantial 1920–21 decrease in the US price level, apparent from Figure 1.7, set the stage for the high real returns on Treasury securities in the 1920s displayed in Figure 1.5.

Foreign Credits

We do not know market values of the foreign credits depicted in Figure 1.3 because these intergovernmental obligations were not marketable.45 But a sequence of defaults, reschedulings, and repudiations eventually drove plausible estimates of the discounted values of prospective payment streams to the US government below the book values of those foreign credits.46

Figure 1.16 reports three time series that summarize original book values (the blue line), renegotiated book values (the red line), and what these payment streams would have been worth if there had been perfect foresight (the green line). Thus, in place of observed market values, the green line shows the present value of what ex post were the continuation flows of actual payments to the Treasury at each date. The blue line shows original book values before a sequence of Treasury write-downs of those book values following a sequence of reschedulings in the early and mid-1920s, recorded as the red line in Figure 1.16. The red line tracks our estimates of what would have been the market values of claims to those promised flows had these flows been risk free and had the claims to them been traded. We formed these estimates by valuing promised flows of payments to the Treasury by the Hicks-Arrow prices of risk-free claims that we inferred from market prices of the Treasury’s own securities.47 Because it uses Hicks-Arrow prices for riskfree claims, the red line undoubtedly overstates what those credits would have traded for had they been marketable. The three largest negotiated settlements are readily apparent: Great Britain, signed on June 19, 1923, and recorded on the Treasury’s books in July 1923; Italy, signed on November 14, 1925, and recorded on the Treasury’s books in May 1927; and France, signed on June 15, 1927, and recorded on the Treasury’s books in April 1930.

Figure 1.16.Face Values and Capitalized Values of Promised and Realized Flows from Foreign Credits

Source: United States World War Foreign Debt Commission (1927), Hall et al. (2020), and author calculations.

Figure 1.17 is a counterpart to Figure 1.6. It shows nominal and real values of $100 invested in the Treasury’s portfolio of foreign credits in December 1919 with earnings being continuously reinvested in the portfolio.48Figures 1.16 and 1.17 confirm that ultimately the foreign credits extended by the US to combatants during World War I failed to provide much “backing” behind the Treasury bonds issued to finance those credits. Ex post, those credits turned out mostly to be subsidies.

Figure 1.17.Nominal and Real Values of $100 Invested in December 1919 in the Foreign Credits Owed to the US Treasury

Source: United States World War Foreign Debt Commission (1927), Hall et al. (2020), and author calculations.

To indicate the burden that the write-downs of these foreign credits imposed on the American taxpayer, we report in Figure 1.18 the annual flows of earnings that would have flowed from the capitalized values in Figure 1.16 as percentages of annual official Treasury interest payments. The blue line reports the implied annual cash flow from the total principal of foreign credits amortized at 5 percent, the original interest rate on Treasury loans to foreign governments. The red line reports the renegotiated promised repayments. Even after the rescheduling, these promised repayments represented roughly 30 percent of officially reported interest payments. Two back-of-the-envelope calculations put magnitudes of the reschedulings into perspective:

  • 1. In 1930 the Allies were scheduled to make $218 million in payments to the US; without the reschedulings, the Allies would have had to pay 5 percent of the outstanding principal, $544 million. This $326 million reduction in payments amounted to $2.67 per person in the US at a time when US per capita income was $748.To translate this reduction from dollars to tax rates, note that in 1929 the Treasury proposed and Congress passed a temporary tax cut49 that reduced tax rates on individual incomes by 1 percentage point50 and reduced the tax rate on corporate income from 12 percent to 11 percent. The Treasury estimated that this tax reduction would “with reasonable accuracy” decrease income tax revenue by $160 million per year, which equaled half the difference between the pre-haircut and post-haircut payment flows that we have estimated.51
  • 2. The difference between the pre-and post-haircut payment flows was about four-tenths of 1 percent of GDP. The Treasury raised about 2 percent of GDP using the personal income tax with total federal revenue being roughly 5 percent of GDP.From 1925 to 1931, the top marginal tax rate was 25 percent. To increase revenue from 2 percent to 2.4 percent of GDP, the Treasury would have had to raise rates by 20 percent (assuming no reduction in the tax base). Assuming a parallel shift in the tax schedule, the top marginal rate would have had to rise to 30 percent.

Figure 1.18.Promised Flows of Payouts from Foreign Credits as Percentages of Official Interest Payments on US Treasury Debt

Source: United States World War Foreign Debt Commission (1927), US Treasury Annual Report of the Secretary of the Treasury on the State of Finances, and author calculations.

Bond Prices and Quantities

This section describes how we have transformed information in US Treasury accounts to match concepts in macroeconomic theory. We describe concepts appearing in Figures 1.3 and 1.4 that report total government debts and in Figures 1.5 and 1.6 that report real and nominal returns on government debt. We link these concepts to decompositions of representations of government budget constraints that appear in government accounts and in macroeconomic theories.

Anne 1.1 describes an accounting framework that reports the relationships among objects in play. Among these objects for each date t are:

  • A list of bonds i = 1,...,nt
  • For each bond i, a price pit, a list of promised future coupons payments ci,t+jt,j=1,...,J(i), and a book value that takes the form of principal payment bi,t+j+J(i)t that the government promises to pay when the bond matures at t + J(i)
  • A list of Hicks-Arrow prices {qt+jt}j=0J(t) telling the number of dollars that at time t exchange for a government promise to pay one dollar at time t + j52
  • A Hicks-Arrow-Debreu pricing equation pit=Σjqt+jtci,t+jt+qi,t+J(i)tbi,t+J(i)t that links the price of each bond at time t to the present value of its coupon stream and principal
  • Sums over bonds of promised coupons ct+jt=Σici,t+1t and sums over bonds of promised principals bt+jt=Σibi,t+jt that, when added, form the stream of payments st+jt=ct+jt+bt+jt,j=1,...,J(t) that at t the government has promised (The notation st+jt is intended to connote “strips.”)
  • A stream {yt} of Treasury net earnings from credits to foreign governments, a stream subject to defaults, renegotiations, repudiations, and extensions of more credits
  • A sequence of estimates {C,} of the discounted present value of the continuation of the earnings stream {yt}

Anne 1.1 uses these concepts to explain discrepancies between objects that in the US Treasury accounts and in macroeconomic models bear the same names. For example, the Treasury measures total government debt by face or book value Σjbt+jt whereas the object in a typical government budget constraint of macroeconomic theory is the market value Σjqt+jtst+jt. Figure 1.4 plots their ratio, which stays as close to unity as it does as a result of a conjunction of the Treasury’s debt management policy—its choice of a division of {st+jt} sequences between {ct+jt}and{bt+jt} sequences—and realized values of the bond yields that represent the Hicks-Arrow vector {qt+jt} at each date in the manner described in Anne 1.1. Figure 1.19 documents that the undiscounted sum of future coupons promised Σjct+jt has sometimes been nearly as large as the sums Σjbt+jt that the Treasury reports as its debt at time t. The Treasury and Federal Reserve jointly decide the division of Treasury obligations to the public between these two sums when they conduct “debt management” and “open market” operations.

US Treasury and macroeconomic theoretic accounts also differ in the quantities to which they attach the word “interest.” These different quantities answer different questions. The Treasury interest concept measures total coupon payments coming due at time t. That quantity helps to inform “cash-management” policy because it estimates a component of the total cash payments that the Treasury is obligated to pay at t. The macroeconomics interest concept is the net rate of return—nominal or real—that the Treasury pays on a value-weighted portfolio of its outstanding bonds. Figure 1.20 compares time series instances of these two concepts. That the nominal return on the government bond portfolio in Figure 1.20 is more volatile than the Treasury’s measure of interest payments reflects fluctuations in market interest rates and capital gains on the Treasury’s bond portfolio that are intermediated through the Hicks-Arrow prices {qt+jt}. For details, please see our discussion of equation (1.1.15) in Anne 1.1.

Figure 1.19.Total Principal Σj=1ntbt+jt and Principal Plus Coupons Σj=1ntbt+jt+Σj=1ntct+jt as Percentages of GDP

Source: Hall et al. (2018),, and author calculations.

Note: Privately held US Treasury debt. It excludes debt held in US government accounts and by the Federal Reserve.

Figure 1.20.Nominal Returns (Thin Black Line) and Official Net Interest Payments (Thick Blue Line) as Percentages of the Par Values of Debt, Annual by Fiscal Year

Source: Hall et al. (2018) and author calculations.

Post–World War I Tax Policy

Congress repeatedly raised income tax rates during World War I, with rates at the top bracket eventually reaching more than 75 percent. Figure 1.21 graphs marginal income tax rates for the years between 1918 and 1925 and shows tax rate reductions after the war. In his first Annual Report in 1921, Secretary of Treasury Andrew Mellon made what later would be called a “supply-side” case for lowering income tax rates. He argued that the high tax rates at the upper bracket discouraged initiative, diverted savings into tax-exempt state and local bonds, and discouraged investors from realizing capital gains. In several subsequent Annual Reports, Mellon observed that increases in top marginal tax rates coincided with decreases in the number individual returns filed that reported incomes over $300,000. Table 1.3, reproduced from Mellon’s 1924 report, shows that, as the marginal tax rate at the top bracket climbed from 15 percent in 1916 to more than 70 percent in 1920, the number of tax returns that reported incomes over $300,000 fell from 1,296 to 395, with a corresponding drop in total income reported at the top bracket from $992 million to $246 million. This occurred even while total personal income rose from $6 billion to $23 billion over this period. Mellon’s supply-side arguments persuaded Congress to cut tax rates. As shown in Figure 1.21, by 1925, Congress had reduced the top marginal income tax rate to 25 percent. The 1925 income tax rate structure allowed the federal government to collect about 2 percent of GDP from the income tax. See Table 1.4.

Figure 1.21.Marginal Income Tax Rates, 1918–25

Sources: The tax schedules are from Tax Foundation (2013). The income thresholds are from Piketty and Saez (2003).

Table 1.3.Tax Returns of Those with Net Income in Excess of $100,000 and $300,000, as Compared with Total of All Net Incomes Returned, for the Calendar Years in Which the Tax Accrues
YearIncome tax, maximum rateTotal amount, of net income returnedNumber of returns, of net income in excess of $100,000Net income returned by those returning in excess of $100,000Percent (5) is of (3)Number of returns of net income in excess of $300,000Net income returned by those returning in excess of $300,000Percent (8) is of (3)
191615 percent$6,298,577,6206,633$1,856,187,71029.471,296$992,972,98615.77
Source: Annual Report of the Secretary of the Treasury, 1924, page 9.
Source: Annual Report of the Secretary of the Treasury, 1924, page 9.
Table 1.4.Revenue from Individual and Corporate Income Taxes
Nominal1929 DollarsPercent of GDPPer Capita
Source: Nominal Revenue (column 2) is from the Annual Report of the Secretary of the Treasury on the State of the Finance, 1930, Table 6, pp. 497–494. The numbers in columns 3–5 are calculated by the authors. GDP, population, and the GDP deflator are from
Source: Nominal Revenue (column 2) is from the Annual Report of the Secretary of the Treasury on the State of the Finance, 1930, Table 6, pp. 497–494. The numbers in columns 3–5 are calculated by the authors. GDP, population, and the GDP deflator are from

The cuts in tax rates appear to have worked as Mellon had reckoned. Figures 1.22, 1.23, and 1.24 report income tax revenues paid by individuals divided into five income groups as percentages of their personal incomes, as totals, and as shares of total income. Between 1918 and 1922, individuals reporting incomes over $100,000 paid an average of 43.1 percent of their income in taxes and 35.5 percent of all income taxes. During the period of the Mellon tax cuts, from 1923 to 1931, individuals reporting incomes over $100,000 paid only 17.1 percent of their income in income taxes, but the income tax revenues that they paid rose from $200 million in 1921 to $450 million in 1927 and to over $700 million in 1928. Further, their share of total income taxes paid rose to 50 percent.

Figure 1.22.Average Tax Rate Paid by Income Groups

Figure 1.23.Individual Income Tax Revenues Paid by Income Groups

Figure 1.24.Shares of Income Taxes Paid by Income Groups

Income Tax Rate Changes Near Start of the Depression

In 1931, Congress raised income tax rates across the board, returning marginal tax rates on upper income taxpayers to pre-1925 levels. Those rate increases helped raise total revenues from the individual income tax by a factor of four from 1931 to 1936. But despite the sharp increase in taxes on the high-income taxpayers, the increased tax burden fell disproportionately on lower income groups; between 1932 and 1939, individuals reporting incomes over $100,000 paid 42.3 percent of their income in taxes, but their share of income taxes paid fell back to 36.6 percent.

Post–World War i Debt Management

Figure 1.3 shows that privately held US Treasury debt peaked in August 1919 at $24.3 billion. By December 1930 it had fallen to $14.2 billion, a reduction of over 40 percent that took place through a steady reduction of nearly $1 billion per year.53Figure 1.25 shows that the average maturity of the federal debt rose during the war, but steadily declined afterward.54

Figure 1.25.Average Maturity of Privately Held US Treasury Debt

Source: Hall et al. (2018) and author calculations.

World War I left Congress a debt management challenge. In October 1919, the US Treasury had $26.2 billion in debts and a repayment schedule that would require it to pay $4.5 billion when the Victory Loans came due in May 1923 (roughly 5 percent of GDP), $4.0 billion in 1928 when the Third Liberty Loan came due, and then relatively little until 1938. Figure 1.26 shows that the four Liberty and Victory Loans composed 80 percent of federal debt in 1920. Thus, financing the war with these medium-term bonds brought “echo effects” when large quantities of debt matured at a small set of dates.55

Figure 1.26.Privately Held US Treasury Debt

Between 1920 and 1928, Secretary Mellon rescheduled the Victory Liberty Loan and Third Liberty Loan by replacing long-term bonds with issues of a set of standardized new short-term securities: term notes and certificates of indebtedness. Mellon also refinanced the Second Liberty Loan. Figure 1.27, which reports our estimates of Treasury strip sequences {st+jt} for two dates shows both the echo effects and the effects of Mellon’s refinancing operations on continuation sequences of strips.

Figure 1.27.Debt Service Profiles, 1920 and 1928

Source: Hall et al. (2018) and author calculations.

Note: “The Mellons” were six long-term bonds issued between 1922 and 1928. Four were used to refinance the Victory Loans and the Second and Third Liberty Loans.

We credit these refinancing operations to Mellon rather than to Congress because World War I brought a significant change in the assignment of authority over debt management. Before World War I, Congress designed each federal bond and typically specified the purposes for which the proceeds could be spent. That first changed with the Second Liberty Bond Act of 1917, which began a 20-year process during which Congress delegated virtually complete authority for bond design to the Treasury. In addition, starting with the Second Liberty Bond Act of 1917, Congress allowed the Treasury to issue bonds not tied to specific projects or designated expenditures.56

Evolution of the Treasury Debt-to-GDP Ratio

We use the government’s consolidated budget constraint to decompose the evolution of the interest-bearing debt-to-GDP ratio into contributions made by nominal returns paid on Treasury securities of different maturities, GDP growth, inflation, Federal Reserve purchases and sales of Treasury securities, and net earning on the Treasury’s foreign credits.57 Using notation presented in Anne 1.1, let Yt be real GDP and vt the real value of a dollar (that is, the inverse of the price level) at time t. Let Bt be the market value of privately held Treasury debt58 and Ct be our estimate of the market value of foreign credits owed to the Treasury. Dividing each side of equation (1.1.19) by nominal GDP, Ytνt, and rearranging terms yields:

where rt1,tj is the nominal return on a j-period zero coupon bond between t –1 and t;rt-1,t denotes the value-weighted net nominal return on Treasury bonds; gt-1,t denotes growth in real GDP, and πt-1,t denotes inflation. The primary deficit, Gt -Tt, is the difference between total government spending and tax revenues. As described in equation (1.1.17) in Anne 1.1, ft is the net one-period payout to the Treasury from t - 1 to t on the Treasury’s portfolio of foreign credits, Ct-1. We let Mt denote the part of high-powered money that is secured by collateral in the form of Treasury securities at time t (please see earlier section titled “Financial Repression and Subsides” and Figures 1.13 and 1.14 again), and we let d denote the discount rate on loans that the Federal Reserve makes to the public for the purpose of purchasing Treasury securities.

Among the terms in equation (1), we have independent measures of the market value of the Treasury debt, Bt; the present value of the promised stream of foreign payments, Ct; government spending and tax collections, GtTt; the stock of Federal Reserve credit secured by Treasury securities, Mt; real GDP, Yt; and the inverse of the price level, vt. Anne 1.1 tells how we computed ft by taking into account the credits extended to the Allies during and immediately after the war and then tracking annual payments and repayments, and also by imputing capital losses to prospective payouts as a result of reschedulings, defaults, and repudiations.59

Data Sources

We used the formula Bt=Σjqt+jtst+jt described in an earlier section to compute the market value of Treasury debt Bt. Data for promised payments streams {st+jt}j at times t are from Hall, Payne, and Sargent (2018). Hall and others (forthcoming) inferred Hicks-Arrow price vectors {qt+jt}j at each date t from prices of individual bonds at each date recorded in Hall, Payne, and Sargent (2018).60 We obtained data on government spending and revenues comprising the “primary” or net-of-interest deficit Gt — Tt from issues of the Annual Report of the Secretary of the Treasury on the State of the Finances. We constructed estimates of the value of foreign credits C by discounting the continuation streams of scheduled payments {yt+j}j. country by country at each date t using the same zero-coupon yield curves that we used to price the Treasury debt.61,62 We took into account changes in the promised flow of payments arising from renegotiations and repudiations. Notably, we assumed that market participants recognized in 1932 that President Hoover’s moratorium on reparations and war debt payments would be permanent for the major debtor countries. Individual country repayment schedules are reported in US World War Foreign Debt Commission (1927). To compute foreign payouts,^ we collected records of the interallied payments from various issues of the Annual Report of the Secretary of the Treasury. Our measure of Federal Reserve Credit Mt is the sum of discounted bills secured by government obligations and government securities owned outright by the Federal Reserve.63 The discount rate dt is the New York Federal Reserve discount rate for loans secured by Liberty and Victory Loans on December 31 of year t.64

For various values of t and τ, Table 1.5 reports our decompositions of debt-to-GDP increments νt(BtCt)Ytνt1(BtτCtτ)Ytτ into components attributable to (1) nominal interest payments, (2) GDP growth, (3) inflation, (4) the primary deficit, (5) net payouts on Treasury-owned foreign credits, (6) Federal Reserve Credit, (7) payments to the Federal Reserve, (8) the cross term, and (9) the residual. Table 1.5 further decomposes contributions of nominal interest payments, GDP growth, and inflation by maturity.

Table 1.5.Contributions to Changes in the Ratio of Debt Net Foreign Credits to GDP
Nominal Return20.130.22-0.514.705.389.2919.21
All maturities
j ≤
2 ≤ j ≤ 40.020.03-0.010.811.022.144.01
j ≥ 50.100.18-0.533.263.926.6713.61
Real GDP Growth2
All maturities net fc-0.32-0.22-0.40-2.53-0.93-11.40-15.81
j ≤ 1-0.01-0.01-0.14-0.66-0.23-1.76-2.81
2 ≤ j ≤ 4-0.02-0.04-0.20-1.26-0.65-3.11-5.27
j ≥ 5-0.30-0.18-0.75-3.41-1.06-5.03-10.72
Foreign credits0.692.791.01-1.502.98
All maturities net fc-0.12-0.42-1.24-0.530.95-2.14-3.49
j ≤ 1-0.00-0.01-0.27-0.110.32-0.01-0.08
2 ≤ j ≤ 4-0.01-0.05-0.42-0.070.36-0.57-0.77
j ≥ 5-0.11-0.36-1.81-0.541.600.03-1.18
Foreign credits1.260.19-1.32-1.58-1.46
Primary Deficit-0.190.9726.22-4.17-10.1319.7332.42
Foreign Payouts-12.78-1.204.9914.045.05
Federal Reserve Credit-0.12-4.252.69-0.66-1.84-4.18
Payments to Federal Reserve-0.04-0.68-0.37-0.33-1.43
Cross Term0.01-0.031.900.470.100.082.52
Note: All contributions are shares of GDP.

Treasury debt is its end-of-year market value net of foreign credits, holdings of the Federal Reserve and government accounts, and the balance in the Treasury.

Treasury debt is decomposed into three groups: claims maturing within one year, j ≤ 1; claims maturing between two and four years, 2 ≤ j ≤ 4; and claims maturing in five years or more j ≥ 5.

Note: All contributions are shares of GDP.

Treasury debt is its end-of-year market value net of foreign credits, holdings of the Federal Reserve and government accounts, and the balance in the Treasury.

Treasury debt is decomposed into three groups: claims maturing within one year, j ≤ 1; claims maturing between two and four years, 2 ≤ j ≤ 4; and claims maturing in five years or more j ≥ 5.

Table 1.5 and Figure 1.28 reveal the following patterns in the ways that the US borrowed, repaid, grew, deflated, inflated, and paid its way toward higher or lower net-debt/GDP ratios:

  • 1. Prior to the US entry into the war, from 1910 to 1916, the net-debt/GDP ratio fell from 3.16 to 2.13. Of this 1.03 percent drop, 0.54 percent was due to real GDP growth and 0.54 percent was due to inflation.
  • 2. During the war period from 1916 to 1918, the increase in the net-debt/GDP ratio from 2.13 to 12.72 was driven by large primary deficits. The extension of foreign credits offset a little less than half of these deficits. Negative nominal returns to bondholders, robust real GDP growth, a burst of inflation, and support from the Federal Reserve also mitigated the impact of these deficits on the net-debt/GDP ratio. The strong GDP growth and high inflation are apparent in Figure 1.29. The negative nominal returns are evident in Figures 1.5 and 1.6.
  • 3. From 1918 to 1931, primary surpluses more than offset high nominal returns to bondholders. The rescheduling of the foreign credits lowered their value so that the contribution of foreign payoffs adds to the net-debt/GDP ratio during this period. In addition, the decline in the price level during the 1920s helped to push up the net-debt/GDP ratio.
  • 4. From 1931 to 1938, the large increase in the net-debt/GDP ratio was driven by large fiscal deficits and the writing off of the foreign credits following the Hoover debt moratorium in 1931. With a few exceptions, payments on Allied war debt to the US stopped in 1933.65

Figure 1.28.Cumulative Sum of the Components of the Change in the Ratio of the Debt Net of Foreign Credits to GDP

Source: See section 10.0.1.

Figure 1.29.Real GDP Growth and Inflation Measured by the GDP Deflator

Evidently, most of the impacts of nominal returns, GDP growth, and inflation on the debt-to-GDP ratio came via their effects on bonds with maturities greater than 5 years. During most of the period under consideration, the average maturity of the debt generally exceeded 8 years (see Figure 1.25) since most the debt consisted of long-term bonds such as the Liberty Loans. See Figures 1.26 and 1.27. Over the entire period from 1910 to 1938, contributions to changes in the debt-to-GDP ratio coming from nominal returns to bondholders were mostly offset by GDP growth and inflation.

Epilogue: More Complications

The earlier section titled “Comparison with Civil War” described how in 1933 President Roosevelt followed Irving Fisher’s advice by adopting policies designed to redistribute wealth from creditors to debtors by increasing the price level. The Roosevelt administration’s efforts in that direction were dwarfed by fiscal-monetary policies that more than a decade earlier had created the German hyperinflation of 1921–23.66 Germany had financed only a small fraction of World War I by borrowing from foreigners; instead, a substantial fraction was financed by Germans who purchased German government securities. By ultimately increasing the price level in November 1923 to approximately 12 orders of magnitude of the 1913 price level (that is, 1012 in scientific notation), the German fiscal-monetary authorities defaulted on virtually all of their domestic debt.67

But the Versailles Treaty in June 1919 and the London Schedule of Payments in April 1921 imposed foreign debts on Germany in the form of uncertain and large reparations payments due in gold to the victors of the war, especially France and the UK. As part of these reparations, Germany was required to pay the Allies a fixed annuity of 2,000 million gold marks ($476 million), roughly the same amount the Allies were scheduled to the United States.68 Although Germany was not required to make reparation payments directly to the US, Germany was required to reimburse the expenses of the American army of occupation and the pay of a set of private claims of American citizens. The German hyperinflation of 1922–23 was an outcome of efforts by reparations creditors, especially France, to force Germany to tax its citizens and suppress its government expenditures to make enough “fiscal space” for Germany to service its reparations obligations. France threatened to occupy the Ruhr Valley and seize factories and mines unless Germany paid reparations payments due in 1922. Germany did not pay. In January 1923, France carried out its threat by invading the Ruhr and then operating Ruhr factories and mines. The German government responded with a “passive resistance” in which it printed German marks to pay German workers not to work for the French. By the fall of 1923, printing presses were financing over 95 percent of the Weimar Republic’s federal expenditures.

The German hyperinflation ended with a bargain among German government authorities, German reparations creditors, the Reparations Commission, and US bankers and government officials that rescheduled German reparations obligations, rearranged German monetary-fiscal institutions and policies in ways designed to protect central bank independence, and brokered an international loan to the German federal government. The Dawes Plan organized a substantial 1924 Reparation Loan to the Reich.69 Under the terms of the Dawes loan, offered by a consortium of banks led by J.P. Morgan & Co., the German Republic borrowed $200 million from private lenders for 25 years at 7 percent; $110,000,000, or roughly half of the total, was sold to US investors. Along with this loan, the Dawes Plan rescheduled reparation payments and reduced them in the short term. The plan placed the Reichsbank under international supervision and German fiscal affairs under the supervision of an Agent General for Reparations of the Allied Reparations Commission.

For five or six years, the Dawes Plan succeeded in easing Germany’s financial distress and converting Germany into an attractive location for foreign investment, particularly from American savers. From 1924 to 1930, Germany became the largest European recipient of American private lending, and the US became Germany’s largest creditor, with the US holding over 40 percent of all German external loans.70 From 1924 to 1930, American investment banks publicly offered 135 dollar-denominated bonds issued by German government entities and dozens of privately offered loans for an aggregate par value exceeding $1.2 billion. Figure 1.30 plots as a blue line the implied quantity outstanding for the 148 bonds listed in Kuczynski (1927, 1932). By June 1931, US banks held $500 million in short-term loans owed by Germans, composing half of all US bank lending to Europe.71 Additional short-term credits, primarily in the form of commercial credits by American firms, brought the total quantity of German debt to US private creditors to over $2,000 million by 1931.

Figure 1.30.Various Estimates of Private US Lending to Germany

Notes and sources: The blue line, labeled “bonds outstanding,” reports the par value outstanding for 135 publicly placed and 13 privately placed dollar-denominated bonds issued in the US and listed in Kuczyns-ki (1927, 1932). For this set of bonds, the quantity outstanding on December 31, 1930, is $1,047 million. However, data from other sources suggest that we have an incomplete list of privately placed bonds; further, we are missing secondary market purchases of German bonds placed in foreign countries.

The series “long-term debt” (red line) and “long-term+short-term debt” (green line) are computed from Table 5, German External Debt, Excluding Reparations, of Klug (1993). We assume that US shares of German external short-term debt and long-term debt remain fixed at 31.8 percent and 49.3 percent, respectively. We convert Reichsmark (RM) into dollars at the rate 0.2382 $/RM.

The O reports quantity outstanding of long-term portfolio investments ($1,117 million) in December 1930. The Δ reports the sum of long-term portfolio and direct investments outstanding ($1,361 million) in December 1930. Both figures are from Table 1 of Klug (1993). See also Dickens (1931, Table 4); Dickens reports that for December 1930 the quantity outstanding of long-term portfolio investments is $1,177 million and the sum of long-term portfolio and direct investments outstanding is $1,421 million.

The * reports US long-term credits outstanding ($1,230 million) in May 1932. The × reports the sum of long-term and short-term credits outstanding ($2,000 million), in May 1932. Both of these figures are reported in Table 2 of Klug (1993).

The ◊ is the quantity outstanding of German bonds issued in the United States, June 1932 ($1,327 million), reported in Table 2 of Klug (1993).

US lenders extended loans to German public and private entities. For 135 publicly offered and 13 privately offered loans listed in Kuczynski (1927, 1932), Table 1.6 divides German borrowers into eight sectors and reports the distribution of loans among them. The only loans to the German Republic (that is, the federal government) were the Dawes loan of 1924 and the Young loan of 1930. Most foreign lending to Germany instead went to Germany’s states, provinces, cities, and industrial firms. Reparations payments were the responsibility of the German Republic. The German Republic encouraged commercial borrowing: after all, if default loomed, the Weimar government could argue that non-reparations public and commercial debts should be paid before reparations. German fiscal authorities hoped to align US private creditors’ interests with theirs and thereby drive a wedge between the interests of US private creditors and reparations creditors. It was not polite to say things like this in public, but German Foreign Minister Gustav Stresemann did. In 1925, he remarked: “One must simply have enough debts; one must have so many debts that, if the debtor collapses, the creditor sees his own existence jeopardized.”72 Of course, that would pit the interests of American private creditors against those of owners of US Treasury debt who had expected the US Treasury’s debt to be backed by its holdings of foreign credits to former World War I allies, which in turn were ultimately backed by German reparations payments.

Table 1.6.Distribution by Sector of German Bonds Issued in the United States, 1924–32
BorrowerNumber of LoansDollar Amount Offered in USPercent of Total
German Republic2$208,250,00016
Provinces, Counties, and19101,420,0008
Public Utilities40257,548,00020
Industrial Corporations35246,968,50019
Credit and Saving Institutions22263,500,00020
Religious and Welfare413,000,0001
Note: This table reports the par value of the loans listed by Kuczynski (1927, 1932) decomposed by sector. Because some of the principal was repaid, this is does represent the total outstanding.
Note: This table reports the par value of the loans listed by Kuczynski (1927, 1932) decomposed by sector. Because some of the principal was repaid, this is does represent the total outstanding.

American isolationists and populists asserted that American advocates of “cancellationist” measures wanted to protect the interests of international bankers and European reparations creditors at the expense of American taxpayers. The New York American criticized “the campaign of international bankers to squeeze, cajole, wheedle Uncle Sam into cancelling the thousands of millions of dollars owed the United States Treasury by European nations” and warned its readers that “international bankers, to protect their own private investments overseas, want the U.S. to cancel the foreign war debts.”73

Were those accusations justified? Not entirely. First, the $2 billion that German borrowers owed US investors was a small fraction of the $12 billion the Allies owed to the US Treasury. Second, American creditors to Germany were also US taxpayers. We have incomplete information about the identities of American creditors, but we do know that most German bonds sold to Americans were issued in small denominations. For example, the Dawes Reparation Loan— officially named the German External Loan, 7 percent Bonds of October 1924— consisted of bonds in denominations of $1,000, $500, and $100, indicating that these bonds were sold to small investors. Using records from 24 American bond houses, J.P. Morgan & Co. partner Dwight W. Morrow (1927) estimated that 91 percent of buyers of Dawes loan bonds purchased less than $5,000 worth and that these small buyers purchased 53.6 percent of the total offering. In addition, in the 1950s, the commission that settled claims of American owners of German dollar-denominated bonds processed 40,620 separate claims.74 This evidence suggests that many retail purchasers of these bonds were individuals rather than banks and other financial firms.

Credits Recognized to Be Subsidies

During the late 1920s, the Federal Reserve raised interest rates in attempts to rein in sharp rises in US stock prices. Higher US interest rates increased the attractiveness of investing in US bonds while also making it more costly for heavily indebted countries like Germany to service their debts to the US. So US foreign lending declined sharply soon thereafter, and international capital flows slowed to a trickle. Germany and several South American countries defaulted. A global financial crisis occurred.

On June 20, 1931, President Hoover proposed a temporary debt moratorium that applied to war credits as well as reparations, but it did not go into effect until the winter when Congress approved it. In December 1932, when the temporary moratorium expired, reparations stopped. Germany stopped making payments on the Dawes and Young loans, although these bonds continued to be traded on US markets well into World War II, albeit at deep discounts.75,76 See Figure 1.31. France repudiated its World War I credits due to the US, but the UK did not, so negotiations with the UK continued, although UK payments to the US had stopped. The US continued to seek compensation for its World War I credits to allies, but only Finland continued to make its full scheduled payments after the moratorium.77

Figure 1.31.US Prices of the Dawes and Young Loans

Source: Wall Street Journal

After all this, between 1918 and 1940, in exchange for the over $12 billion in loans, credits, and accrued interest on the US Treasury’s books, the Treasury received $2.86 billion in remittances ($2.11 billion labeled interest and $777 million labeled principal) from the Allies. The UK made $2.19 billion in payments ($464 million in principal and $1.72 million in interest) on the $4.7 billion the UK government owed the US Treasury, whereas France paid only $486 million ($226 in principal and $260 million in interest) on the $4.2 billion that the government of France owed the US Treasury.

Annex 1.1. Accounting Framework

This annex compares accounting systems used by the US Treasury and macroeconomic theory.

Government Budget Constraint

We want to represent the government budget constraint as it appears in macroeconomic models. Let Bt1=Σj=1nBt1j be the total nominal value of interest bearing government debt at t-1, where Bt1j is the nominal value of zero coupon bonds of maturity j at t-1. Let rt1,tj be the net nominal return between t-1 and t on nominal zero coupon bonds of maturity j. The government budget constraint at time t is


where Gt is the dollar value of government purchases, Tt is the dollar value of taxes net of transfers, Mt is the stock of non-interest-bearing government debt called base money, and the equality

implicitly defines the value-weighted net nominal return rt-1,t on interest-bearing nominal government bonds from t-1 to t. In a later section of this annex titled “Foreign Credits and Federal Reserve Lending,” we modify equation (1.1.2) to account for foreign credits.

Let Yt be real GDP at t, pt be the price level, νt=pt1 be the real value of a dollar, and vtBt be the real value of interest-bearing government debt to the public. The government budget constraint equation (1.1.2) or (1.1.3) and simple algebra tell how a net nominal return rt-1,t , a net inflation rateπt-1,t, a net growth rate in real GDP gt-1,t, a net rate of increase in base money µt-1,t, and a real primary deficit deft = vt (Gt -Tt) contribute to the evolution of the government interest-bearing debt-to-GDP ratio:

To bring out consequences of interest rate risk and the maturity structure of the debt for the evolution of the debt-to-GDP ratio, we refine equation (1.1.5) to recognize that the government pays different nominal one-period returns on the dollar-denominated IOUs of different maturities that comprise Bt:

Equation (1.1.6) distinguishes contributions to the growth of the debt-to-GDP ratio that depend on debt maturity j from those that do not: πt-1,t and gt-1,t do not depend on j and operate on the total real value of debt last period, but the nominal returns rt1,tj depend on maturity j and operate on the real values of the corresponding maturity j components Bt1j.

US Government Accounts

The Treasury measures government debt and interest payments differently than do macroeconomists. The official accounts measure government debt by the total par value of outstanding promises, while the macroeconomist’s budget constraint is cast in terms of market values.

To understand how the Treasury’s measure of government debt is related to the market value of debt, we bring in information about bonds’ coupons, principals or par values, and prices of (presumably risk-free) promises to future dollars. In the tradition of macro-finance, we use Hicks-Arrow prices of future-dated claims. Let time be discrete so that t ∈ {0,±1,±2,....}. Let the market price qt+jt be the number of dollars at time t that buys a risk-free claim to a dollar at time t+j. The superscript t denotes the date at which the price is quoted, whereas the subscript t+j refers to the date at which a promise to pay is to be fulfilled. At any date t, let there be a list of market prices {qt+jt}j=0nt, where nt is the maximum horizon over which the government has promised payments.78 (Hall and others [forthcoming]) describe how they inferred the Hicks-Arrow prices that we use from a collection of bond prices and associated promised payment streams.79) We set qtt=1 to express that a dollar today costs one dollar today.80 For j ≥ 1, the price qt+jt is related to the yield to maturity ρjt for j-period risk-free zero coupon bonds by

The gross nominal return on a j-period zero coupon bond from time t to t+1 is

where rt,t+1j is the net nominal return. The net return equals the yield only for j = 1.

At time t the government promises to pay st+jt dollars at times t+j, j = 1, 2,...nt.

We interpret stt as currency or base money.81 For j ≥ 1, promised payments consist

of coupons ct+jt and terminal or principal payments (also known as par values) bt+jt:

These are sums over all outstanding bonds of coupon and principal components associated with each bond. The market value of interest-bearing government debt at time t is

which states that the total value of government debt is the sum of a collection of prices times quantities.82

The Treasury defines government debt at time t as the sum of par values of outstanding debt

which differs from the market value of government debt

for two reasons:

  • It neglects promises to pay coupons
  • and
  • The book value given by equation (1.1.9) fails to multiply future principal payments of bt+jt by multiplying them by market prices qt+jt.

The first omission causes the official Treasury concept (1.1.9) to understate the market value of debt, while the second omission makes it overstate it. This means that the official measure of government debt V Σj=1ntbt+jt can either exceed or fall short of the market value Σj=1ntqt+jtst+jt.

We can represent the macroeconomists budget constraint (1.1.2) or (1.1.3) as83

The left side of equation (1.1.11) is the value of government debt in period t, whereas the first term on the right side is the value of payments that the government had promised at time t – 1 evaluated at time time t prices qt+jt. Equation (1.1.11) states that the value of the government debt changes between times t and t+1 because

  • 1. Prices of time t+j promises st+jt1 to time t+j dollars change from qt+jt1toqt+jt.
  • 2. The government pays off or reschedules some components of its promised payments, contributing to deviations of st+jtfromst+jt1 for some j’s .
  • 3. The government runs a net-of-interest nominal deficit or surplus at date t. It is enlightening to rewrite equation (1.1.11) as

The second term on the right side of the second line of equation (1.1.12) measures time t nominal net returns on the time t -1 nominal government debt:

So equation (1.1.12) expresses the nominal value of government debt in period t as the sum of the value of government debt yesterday, net nominal returns on last period’s debt, and the government deficit Gt -Tt.

Interest Reported by the Treasury

The net nominal interest payments defined in equation (1.1.13) are not what the US Treasury reports. Instead, it reports a different notion of interest, namely:

  • 1. Before 1929:
  • 2. After 1929:
  • where b1,tt1 is the par value of pure discount one-period treasury bills issued at t – 1. So what the Treasury reports as interest consists of coupons on longer maturity bonds plus the net yield on one-period zero coupon Treasury bills (these have existed only since 1929).

To relate the government’s post-1929 definition of nominal interest payments to the theoretical concepts in a standard macroeconomic formulation such as equation (1.1.11), we first introduce the decomposition btt1=b1,tt1+b1,tt1whereb1,tt1 is the par (or principal) value of bonds with initial maturities exceeding one period that fall due at time t. (Here we follow game theorists in using the subscript – 1 to mean “not 1,” which in the present context means “not a treasury bill.”) We use this decomposition to accommodate how the US Treasury accounts for interest on Treasury bills. Then note that qtt=1 and rewrite the standard macroeconomic government budget constraint (1.1.12) as

The second and third terms on the second line of the right side of equation (1.1.15) decompose principal payments into those attributable to maturing one-period pure discount bonds b1,tt1 and to maturing longer term bonds b1,tt1. Rewrite the right side of equation (1.1.15) as Σj=2nt1qt+j1t1st+j1t1+(GtTt) plus

The first term is what the Treasury records as interest payments. The second term constitutes repayments of principal at time t. We can interpret the sum of the first and second terms of the above sum as expressing cash that the Treasury must have at time t in order to “service” its debt, meaning to pay coupons plus principal due at time t.84 The third term measures capital gains or losses on longer term government debt held from t – 1 to t. These capital gains are included in the macroeconomic concept of interest on the government debt but are neglected in the official concept.

Foreign Credits and Federal Reserve Lending

The value of the promised stream of payments from foreign governments to the US Treasury, {yt+jt}, at time t is

where qt+jt is the time t Hicks-Arrow price of a risk-free claim to one dollar at time t+j, and mt is the maximum horizon over which the Treasury is promised payments at time t.85 The promised payment stream, {yt+jt}, depends on the history of foreign credits extended at times t and earlier in a way that we now describe.

For convenience and without loss of generality, set mt = +∞ for all t. Let ft be the one-period net payout from the portfolio of credits Ct-1 from t – 1 to t, including the time t coupon or repayment and capital gains or losses from revaluations and reschedulings:

Adding and subtracting Σj=1qt+jtyt+jt1 to the right side of the above equation gives

The first term in equation (1.1.17) for ft is the time t payoff. The second term is the gain from restructuring the payment schedule between t – 1 and t. The third term is the capital gain on the time t – 1 promised payment stream due to the change in Hicks-Arrow prices between t – 1 and t. The initial condition for foreign credits is

which states that at time t = 0 the US government lends – f0 to foreign governments in exchange for a promised repayment stream {yj0}j=1 from foreign governments to the US.

To account for the interest payments on the Federal Reserve’s loans to the public for the purpose of purchasing Treasury securities, we let dt denote the end-of-period discount rate on these loans and let Mt denote the high-powered money secured by collateral in the form of Treasury securities at end of period t. Thus, when the Treasury owns foreign credits and the Federal Reserve extends credit to the public, we modify equation (1.1.2) to be

Equation (1.1.19) indicates how lower income from government foreign credits increases the amount that the Treasury must finance.

In the government accounts, the extension of foreign credits is included in government spending. Thus, if in period t, the government borrows $1 and lends it to a foreign government, Gt , ft , Bt , and Ct each increase by $1.

Annex 1.2. Secretary Bryan’s Letter to President Wilson

I beg to communicate to you an important matter which has come before the Department. Morgan Company of New York have asked whether there would be any objection to their making a loan to the French Government and also the Rothschilds – I suppose that is intended for the French Government. I have conferred with Mr. Lansing and he knows of no legal objection to financing this loan, but I have suggested to him the advisability of presenting to you an aspect of the case which is not legal but I believe to be consistent with our attitude in international matters. It is whether it would be advisable for this Government to take the position that it will not approve of any loan to a belligerent nation. The reasons that I would give in support of this proposition are:

First: Money is the worst of all contrabands because it commands everything else. The question of making loans contraband by international agreement has been discussed, but no action has been taken. I know of nothing that would do more to prevent war than an international agreement that neutral nations would not loan to belligerents. While such an agreement would be of great advantage, could we not by our example hasten the reaching of such an agreement? We are the one great nation which is not involved, and our refusal to loan to any belligerent would naturally tend to hasten a conclusion of the war. We are responsible for the use of our influence through example, and as we cannot tell what we can do until we try, the only way of testing our influence is to set the example and observe its effect. This is the fundamental reason in support of the suggestion submitted.

Second: There is a special and local reason, it seems to me, why this course would be advisable. Mr. Lansing observed in the discussion of the subject that a loan would be taken by those in sympathy with the country in whose behalf the loan was negotiated. If we approved of a loan to France we could not, of course, object to a loan to Great Britain, Germany, Russia, or to any other country, and if loans were made to these countries, our citizens would be divided into groups, each group loaning money to the country which it favors and this money could not be furnished without expressions of sympathy. These expressions of sympathy are disturbing enough when they do not rest upon pecuniary interests – they would be still more disturbing if each group was pecuniarily interested in the success of the nation to whom its members had loaned money.

Third: The powerful financial interests which would be connected with these loans would be tempted to use their influence through the newspapers to support the interests of the Government to which they had loaned because the value of the security would be directly affected by the result of the war. We would thus find our newspapers violently arrayed on one side or the other, each paper supporting a financial group and pecuniary interest. All of this influence would make it all the more difficult for us to maintain neutrality as our action on various questions that would arise would affect one side or the other and powerful financial interests would be thrown into the balance.... As we cannot prevent American citizens going abroad at their own risk, so we cannot prevent dollars going abroad at the risk of the owners, but the influence of the Government is used to prevent American citizens from doing this. Would the Government not be justified in using its influence against the enlistment of the nation’s dollars in a foreign war?

Secretary of State William Jennings Bryan to President Woodrow Wilson,

August 10, 1914

Annex 1.3. US Treasury Credits: Original Principal and Present Value of Refinanced Promised Payments
CountryOriginal PrincipalFunded InterestFunded DebtDebt prior to Funding Including Accrued InterestPresent Value of Refinanced Promised Payments and Present Value as a Percent of Principal prior to Refunding
3 Percent4¼ Percent5 Percent
Great Britain4,074,818,358525,181,6424,175,310,0004,715,310,0004,922,702,000104.43,788,470,00080.33,296,948,00069.9
Total$9,811,094,094$1,711,259,906$11,522,354,000$12,036,376,000$9,175655,00076.2 percent$6,862,285,00057.0 percent$5,873,638,00048.8%
Annex 1.4. franklin D. Roosevelt’s “Bombshell Message”

I would regard it as a catastrophe amounting to a world tragedy if the great Conference of Nations, called to bring about a more real and permanent financial stability and a greater prosperity to the masses of all Nations, should, in advance of any serious effort to consider these broader problems, allow itself to be diverted by the proposal of a purely artificial and temporary experiment affecting the monetary exchange of a few Nations only. Such action, such diversion, shows a singular lack of proportion and a failure to remember the larger purposes for which the Economic Conference originally was called together.

I do not relish the thought that insistence on such action should be made an excuse for the continuance of the basic economic errors that underlie so much of the present world-wide depression.

The world will not long be lulled by the specious fallacy of achieving a temporary and probably an artificial stability in foreign exchange on the part of a few large countries only.

The sound internal economic system of a Nation is a greater factor in its well-being than the price of its currency in changing terms of the currencies of other Nations.

It is for this reason that reduced cost of Government, adequate Government income, and ability to service Government debts are all so important to ultimate stability. So too, old fetishes of so-called international bankers are being replaced by efforts to plan national currencies with the objective of giving to those currencies a continuing purchasing power which does not greatly vary in terms of the commodities and need of modern civilization. Let me be frank in saying that the United States seeks the kind of dollar which a generation hence will have the same purchasing and debt-paying power as the dollar value we hope to attain in the near future. That objective means more to the good of other Nations than a fixed ratio for a month or two in terms of the pound or franc.

Our broad purpose is the permanent stabilization of every Nation’s currency. Gold or gold and silver can well continue to be a metallic reserve behind currencies, but this is not the time to dissipate gold reserves. When the world works out concerted policies in the majority of Nations to produce balanced budgets and living within their means, then we can properly discuss a better distribution of the world’s gold and silver supply to act as a reserve base of national currencies. Restoration of world trade is an important factor, both in the means and in the result. Here also temporary exchange fixing is not the true answer. We must rather mitigate existing embargoes to make easier the exchange of products which one Nation has and the other Nation has not.

The Conference was called to better and perhaps to cure fundamental economic ills. It must not be diverted from that effort.

Franklin D. Roosevelt, Wireless to the London Conference, July 3, 1933


    AdamsHenryCarter. 1887. Public Debts: An Essay in the Science of Finance. New York: D. Appleton and Co.

    AguiarMarkManuelAmadorHugoHopenhayn andIvanWerning. 2016. “Take the Short Route: Equilibrium Default and Debt Maturity.” NBER Working Paper 22847National Bureau of Economic ResearchCambridge, MA.

    • Search Google Scholar
    • Export Citation

    BarroRobertJ.1979. “On the Determination of the Public Debt.” Journal of Political Economy87 (5): 94071.

    BassettoMarco andCarloGalli. 2017. “Is Inflation Default? The Role of Information in Debt Crises.” Working Paper 2017–06 Federal Reserve Bank of Chicago University College London and Institute for Fiscal Studies.

    • Search Google Scholar
    • Export Citation

    BhandariAnmolDavidEvansMikhailGolosov andThomasSargent. 2017a. “Fiscal Policy and Debt Management with Incomplete Markets.” Working PaperUniversity of MinnesotaMinneapolis.

    • Search Google Scholar
    • Export Citation

    BhandariAnmolDavidEvansMikhailGolosov andThomasSargent. 2017b. “The Optimal Maturity of Government Debt.” Working PaperUniversity of MinnesotaMinneapolis.

    • Search Google Scholar
    • Export Citation

    Board of Governors of the Federal Reserve System. 1915–40. Federal Reserve Bulletin. Washington, DC: US Government Printing Office.

    Board of Governors of the Federal Reserve System. 1922. Discount Rates of the Federal Reserve Banks 1914–1921. Washington, DC: US Government Printing Office.

    • Search Google Scholar
    • Export Citation

    Board of Governors of the Federal Reserve System. 1943. Banking and Monetary Statistics 1914–1941. Washington, DC: US Government Printing Office.

    • Search Google Scholar
    • Export Citation

    BuchananJamesM. andT.Nicolaus Tideman. 1975. “Gold, Money, and the Law: The Limits of Governmental Monetary Authority.” In Gold Money and Law edited by Henry G.Manne andRogerLeRoy Miller970. Chicago: Aldine Publishing Co.

    • Search Google Scholar
    • Export Citation

    ClarkePeter. 2017. The Locomotive of War: Money Empire Power and Gold. London: Bloomsbury Publishing.

    DavisJosephH.2004. An Annual Index of U.S. Industrial Production, 1790–1915. The Quarterly Journal of Economics119 (4): 1177215.

    DeweyDavisR.1912. Financial History of the United States. 4th ed. New York: Longmans, Green, and Company.

    DiasDanielA.ChristineRichmond andMark L. J.Wright. 2014. The stock of external sovereign debt: Can we take the data at face value?Journal of International Economics94 (1): 117.

    • Search Google Scholar
    • Export Citation

    DickensPaul. 1931. “A New Estimate of American Investments Abroad.” Tech. Rep. Trade Information Bulletin No. 75US Department of CommerceWashington, DC.

    • Search Google Scholar
    • Export Citation

    EdwardsSebastian. 2018. American Default: The Untold Story of FDR the Supreme Court and the Battle over Gold. Princeton, NJ: Princeton University Press.

    • Search Google Scholar
    • Export Citation

    EichengreenBarry. 1992. Golden Fetters: The Gold Standard and the Great Depression 1919–1939. New York: Oxford University Press.

    FaragliaElisaAlbertMarcetRigasOikonomou andAndrewScott. 2014. “Government Debt Management: The Long and the Short of It.” CEPR Discussion Paper 10281Centre for Economic Policy ResearchLondon.

    • Search Google Scholar
    • Export Citation

    FergusonThomas. 1989. “Industrial Conflict and the Coming of the New Deal: The Triumph of Multinational Liberalism in America.” In The Rise and Fall of the New Deal Order: 1930–1980 edited by StevenFraser andGaryGerstle331. Princeton, NJ: Princeton University Press.

    • Search Google Scholar
    • Export Citation

    FisherIrving. 1933. “The Debt-Deflation Theory of Great Depressions.” Econometrica1 (4): 33757.

    FriedmanMilton. 1975. “Comment on Buchanan and Tideman’s Paper.” In Gold Money and Law edited by Henry G.Manne andRogerLeRoy Miller. Chicago: Aldine Publishing Co.

    • Search Google Scholar
    • Export Citation

    FriedmanMilton andAnnaJacobson Schwartz. 1963. A Monetary History of the United States 1867–1960. Princeton, NJ: Princeton University Press.

    • Search Google Scholar
    • Export Citation

    GallatinAlbert. 1837. “Report on the Finances November, 1807.” Reports of the Secretary of the Treasury of the United States. Vol 1. Washington, DC: US Department of the Treasury.

    • Search Google Scholar
    • Export Citation

    GarbadeKenneth. 2012. Birth of a Market: The U.S. Treasury Securities Market from the Great War to the Great Depression. Cambridge, MA: MIT Press.

    • Search Google Scholar
    • Export Citation

    German Registration Office for Foreign Debt. 1932–40. Die Auslandsverschuldung Deutschlands. Berlin SW III.

    GomesJoaoUrbanJermann andLukasSchmid. 2016. “Sticky Leverage.” American Economic Review106 (12): 380028.

    GuinnaneTimothyW.2004. “Financial Vergangenheitsbewltigung: The 1953 London Debt Agreement.” Working Paper 880Economic Growth Center, Yale UniversityNew Haven, CT.

    • Search Google Scholar
    • Export Citation

    HallGeorgeJ.JonathanPayne andThomas J.Sargent. 2018. “US Federal Bonds 1776–1940: Quantities and Prices.” Jupyter notebooks Python pickel files and Pandas DataFrames.

    • Search Google Scholar
    • Export Citation

    HallGeorgeJ.Jonathan E.PayneThomas J.Sargent andBalintSzoke. Forthcoming. “Approximating US Bond Yields 1791–1940.” New York UniversityNew York.

    • Search Google Scholar
    • Export Citation

    HallGeorgeJ. andThomas J.Sargent. 2011. “Interest Rate Risk and Other Determinants of Post-WORLD WAR II US Government Debt/GDP Dynamics.” American Economic Journal: Macroeconomics3 (3): 192214.

    • Search Google Scholar
    • Export Citation

    HallGeorgeJ. andThomas J.Sargent. 2014. “Fiscal Discriminations in Three Wars.” Journal of Monetary Economics61 (C): 14866.

    HallGeorgeJ. andThomas J.Sargent. 2015. “A History of U.S. Debt Limits.” NBER Working Paper 21799National Bureau of Economic ResearchCambridge, MA.

    • Search Google Scholar
    • Export Citation

    HartmanDouglasW. andWaltherSkaupy. 1957. “Third Annual Report of the Validation Board for German Dollar Bonds.” Department of State BulletinXXXVI (925): 44756.

    • Search Google Scholar
    • Export Citation

    HawtreyRalphG.1919. Currency and Credit. London: Longmans, Green, and Co.

    HilscherJensAlonRaviv andRicardoReis. 2017. “Inflating Away the Public Debt? An Empirical Assessment.” Working PaperLondon School of EconomicsLondon.

    • Search Google Scholar
    • Export Citation

    HofstadterRichard. 1948. The American Political Tradition: And the Men Who Made It. New York: A. A. Knopf.

    International Monetary Fund (IMF). 2014. Fiscal Monitor: Public Expenditure Reform: Making Difficult Choices. Washington, DC.

    KeynesJohnMaynard. 1920. The Economic Consequences of the Peace. New York: Harcourt Brace.

    KeynesJohnMaynard. 1922. A Revision of the Treaty. London: Macmillan and Co.

    KlugAdam. 1993. “The German Buybacks, 1932–1939: A Cure for Overhang?Technical Report 75Princeton Studies in International FinancePrinceton, NJ.

    • Search Google Scholar
    • Export Citation

    KuczynskiR. R.1927. American Loans to Germany. Investigations in International Economic italic Reconstruction. Macmillan.

    KuczynskiR. R.1932. Bankers’ Profits from German Loans. Washington, DC: The Brookings Institution.

    LinkArthurS.1965. Wilson: Campaigns for Progressivism and Peace 1916–1917 Volume V. Princeton, NJ: Princeton University Press.

    LjungqvistLars andThomas J.Sargent. 2018. Recursive Macroeconomic Theory. 4th ed. Cambridge, MA: MIT Press.

    McAdooWilliamG.1931. Crowded Years: The Reminiscences of William G. McAdoo. New York: Houghton Mifflin Company.

    MellonAndrew. 1926. Annual Report of the Secretary of the Treasury on the State of the Finances. Washington, DC: US Department of the Treasury.

    • Search Google Scholar
    • Export Citation

    MeyerG. J.2006. A World Undone: The Story of the Great War 1914 to 1918. New York: Random House.

    MironJeffreyA. andChristina D.Romer. 1990. “A New Monthly Index of Industrial Production, 1884–1940.” The Journal of Economic History50 (2): 32137.

    • Search Google Scholar
    • Export Citation

    MorrowDwightW.1927. “Who Buys Foreign Bonds?Foreign Affairs5 (2): 21932.

    PikettyThomas andEmmanuelSaez. 2003. “Income Inequality in the United States, 1913–1998.” The Quarterly Journal of Economics118 (1): 139.

    • Search Google Scholar
    • Export Citation

    RicardoDavid. 1816. Proposals for an Economical and Secure Currency. London: John Murray Albermarle-Street. Reprinted in Vol. IV of The Works and Correspondence of David Ricardo edited by Pierro Sraffa with the collaboration of M.H. Dobb Liberty Fund Indianapolis 2004.

    • Search Google Scholar
    • Export Citation

    RothbardMurrayN.2002. A History of Money and Banking in the United States: The Colonial Era to World War II. Auburn, AL: Ludwig von Mises Institute.

    • Search Google Scholar
    • Export Citation

    SargentThomasJ.1982. “The Ends of Four Big Inflations.” In Inflation: Causes and Effects4198. Cambridge, MA: National Bureau of Economic Research.

    • Search Google Scholar
    • Export Citation

    SchukerStephanA.1988. “American Reparations to Germany, 1919–1933: Implications for the World Debt Crisis.” In Princeton Studies in International Finance NO. 61. Princeton, NJ: Princeton University.

    • Search Google Scholar
    • Export Citation

    SkidelskyRobert. 1983. John Maynard Keynes Volume One: Hopes Betrayed 1883–1920. New York: Elisabeth Sifton/Viking.

    SteinerZara. 2005. The Lights That Failed: Euopean International History 1919–1933. Oxford, UK: Oxford University Press.

    Tax Foundation. 2013. “U.S. Federal Individual Income Tax Rates History 1862–2013 (Nominal and Inflation-Adjusted Brackets).”–2013-nominal-and-inflation-adjusted-brackets.

    • Search Google Scholar
    • Export Citation

    TaylorFrederick. 2013. The Downfall of Money: Germany’s Hyperinflation and the Destruction of the Middle Class. London: Bloomsbury Press.

    • Search Google Scholar
    • Export Citation

    ToozeAdam. 2006. The Wages of Destruction: The Making and Breaking of the Nazi Economy. New York: Penguin Books.

    ToozeAdam. 2014. The Deluge: The Great War America and the Remaking of the Global Order. New York: Viking.

    United Nations. 1959. “Agreement on German External Debts Signed at London, on 27 February 1953.” Technical Report. 4764 United Nations Treaties and International Agreements New York.

    • Search Google Scholar
    • Export Citation

    US Internal Revenue Service. 1920–40. Statistics of Income. Washington, DC: US Government Printing Office.

    US World War Foreign Debt Commission. 1927. Combined Annual Reports of the World War Foreign Debt Commission with Additional Information Regarding Foreign Debts due the United States. Washington, DC: US Government Printing Office.

    • Search Google Scholar
    • Export Citation

    VinerJacob. 1933. Balanced Deflation Inflation or More Depression. Minneapolis: University of Minnesota Press.

    WoodJohn. 2009. “The Great Deflation of 1929–33 (Almost) Had to Happen.” Department of EconomicsWake Forest UniversityWinston-Salem, NC.

    • Search Google Scholar
    • Export Citation
1We thank William Berkley for supporting our research. Hall thanks the Teodore and Jane Norman Fund for financial support. We thank Mark De Broeck, Dominique Guillaume, Paul Jankowski, Juna Luzi, Jonathan Payne, Jeremy Siegel, Balint Szoke, and Ellis Tallman for helpful comments and Jefery Cheng and Rahim Damji for excellent research assistance. Balint Szoke wrote Python code to estimate the term structure of yields on Treasury bonds, and Jonathan Payne wrote Python code to organize the bond price and quantity data.
2As Hofstadter (1948, ch. 10, Part IV) says of Wilson’s recommendation, “but he and his most important advisers were utterly incapable of obeying the injunction themselves.”
3See Link (1965, 202), Meyer (2006, 422), and Tooze (2014, 51).
4See Skidelsky (1983). President Wilson coined the phrase “peace without victory” in his January 22, 1917, speech to the US Senate. That speech contains a description of conditions for a sustainable peace that President Wilson and the other victors would ignore at the Peace Conference at Versailles: “Victory would mean peace forced upon the loser, a victor’s terms imposed upon the vanquished. It would be accepted in humiliation, under duress, at an intolerable sacrifce, and would leave a sting a resentment, a bitter memory upon which terms of peace would rest, not permanently, but only as upon quicksand. Only a peace between equals can last. Only a peace the very principle of which is equality and a common participation in a common benefit. The right state of mind, the right feeling between nations, is as necessary for a lasting peace as is the just settlement of vexed questions of territory or of racial and national allegiance.” Wilson changed his mind about these things after the US entered the war.
5Hofstadter (1948, ch. 10) and Tooze (2014) document this self-confrming alignment of beliefs.
6Schuker (1988) describes complications and consequences.
7In June 1918, roughly $1.75 billion in private loans to Britain and France remained outstanding.
8The table omits an important feature of Liberty Bonds that underlies the story told by Edwards (2018) : Liberty bonds contained a gold clause promising to pay interest and principal in gold coin. Congress abrogated the gold clause in both private and public debt contracts in 1933, provoking a sequence of legal challenges culminating in four cases decided by the Supreme Court mostly but not entirely in the government’s favor in February 1935. One of the four cases, Perry v. United States, involved Perry’s request that the government honor its promise to pay his Fourth Liberty Loan bonds in US gold coin or its equivalent.
9The conversion rate was lower for the First Liberty Bonds because conversion to a higher coupon bond also meant a less favorable tax treatment.
10Through the Liberty Loan Acts, credits were granted to Belgium, Cuba, Czechoslovakia, France, Great Britain, Greece, Italy, Liberia, Romania, Russia, and Serbia.
11The nine countries granted just postwar credits were Armenia, Austria, Estonia, Finland, Hungary, Latvia, Lithuania, Nicaragua, and Poland.
12For us, the terms face value and par value mean the same thing.
13This question and the previous ones underlie the struggles chronicled by Edwards (2018).
14See the speech by Karl Helfrerich cited by Taylor (2013, ch. 2, Loser Pays All).
15See Keynes (1920, 1922), Steiner (2005, ch. 4), Tooze (2014, 295–99), and Clarke (2017, chs. 11–13).
16The phrases in quotes are again from Adams (1887, 37–38).
17We use some of the data on prices and quantities of all bonds issued by the US government from 1776 to 1960 that have been collected and organized by Hall, Payne, and Sargent (2018).
18Our accounting exercises are designed to shed light on just some of the broader web of complications described with authority and wide command of primary sources in books by Eichengreen (1992), Steiner (2005), and Tooze (2014).
19President Wilson’s Treasury Secretary and son-in-law William Gibbs McAdoo sought to borrow in a more orderly way than he thought the Union had managed during the Civil War. McAdoo (1931, 373) said that he “did not get much in the way of inspiration or suggestion from a study of the Civil War, except a pretty clear idea of what not to do.” For example, rather than improvising 19 separate securities as the Union had during the Civil War, McAdoo convinced Congress to finance the war mostly by sequentially issuing five securities.
20Hall and Sargent (2014) compare fiscal aspects of the Civil War with those of the War of 1812 and the War for Independence.
21As part of his critical analysis of the way the Union had financed the Civil War, Adams (1887, 134–35) observed that “it is easier to raise the rate of existing taxes than to establish a new system of duties. From this it follows that the germ of a war policy lies back in the treasury policy of ordinary times.” He recommended that “the permanent system should be so adjusted as to respond quickly to any change in rates imposed, and this can be easily done by fixing the ordinary rate of taxation below the maximum revenue rate.” Adams (1887, 132) performed an interesting counter-factual experiment in which he estimated the consequences for Union finances of having immediate access to revenue sources that the Congress established only during the war.
22Bassetto and Galli (2017) provide an information theoretic answer to the question “Is Inflation Default?”
23However, during the war, a sizeable fraction of the interest-bearing federal debt was owned by the Federal Reserve System. Please see the third paragraph of later section titled “Financial Repression and Subsidies.”
24President Wilson called them associates, not allies.
25Distinguishing between book and market values of government debt is a theme of Dias, Richmond, and Wright (2014) and Hall and Sargent (2011).
26These experiments were designed to reap the gains in efficiency and price level stability promised by David Ricardo’s 1816 proposal for a well-managed paper currency; see Ricardo (1816, 35).
27Rothbard (2002) contains an imaginative account of these experiments.
28The US embargoed exports of gold for much of 1914, 1918, and 1919.
29The nominal value at time t is 100×Πs=1917:1t(1+rs,s+1), where rs,s+1, is the nominal net return on the portfolio between month s and s + 1. The real value at time t is 100×Πs=1917:1t1+rs,s+11+πs,s+1, where πs,s+1 is the inflation rate between months s and s + 1. Thus, the real value is reported in units of January 1917 dollars.
30See footnote 21.
32Milton Friedman (1975, 75) argued that “departing from gold was not necessary” for Roosevelt to pursue the price level he sought through monetary policy. Friedman noted that Viner (1933) had argued that the US could engineer inflation without raising the price of gold above $20.67 an ounce. See Buchanan and Tideman (1975) for alternative views about Roosevelt’s actions about the gold standard. See Edwards (2018, 39) for an account of a proposal authored by James Warburg to reduce the Federal Reserve’s gold coverage ratio in a way that would allow the Federal Reserve to conduct operations that would facilitate monetary expansion and a controlled rise in the US price level while remaining on gold at the pre-1933 price of 20.67 dollars per troy ounce.
33Again, the US embargoed exports of gold for much of 1914, 1918, and 1919.
34Hawtrey (1919) described the workings of this mechanism during monetary disturbances in France and the UK from 1789 to 1821.
35See footnote 31 for an explanation of how we computed the real value of the post–World War I portfolio. The real value of $100 invested in January 1861 is computed in a like manner and is reported in January 1861 dollars.
36That was the policy advocated by the Committee for the Nation to Rebuild Prices and Purchasing Power that infuential businessmen organized in January 1933. See Rothbard (2002, 297). Rothbard tells how Irving Fisher transferred residual monies to the Committee for the Nation from the defunct Stable Money League that he had formed in 1921 to promote price level targeting.
37Notice Roosevelt’s “specious fallacy” pun about commodity (specie) standards. Notice also how Roosevelt distinguished temporary from permanent policy actions, a distinction that 45 years later would become an important aspect of rational expectations macroeconomics. Rothbard (2002, 307) and Ferguson (1989, 28–29) detected advisers who helped Roosevelt to formulate his policy and to write the “bombshell message” to the London Conference explaining it.
38Gomes, Jermann, and Schmid (2016) create a quantitative structure that extends and formalizes a debt-deflation theory.
39We have not located a source for the fraction of post–Civil War corporate debt that was denominated in gold. We hope to collect that information in the near future.
40See Rothbard (2002, 297–98). Also, see Edwards (2018), especially his appendix on George F. Warren versus Irving Fisher, who used distinct theoretical and empirical justifications and implementation strategies for their price level targeting recommendations.
41Real returns on US government debt between 1933 and 1935 displayed in Figures 1.6 and 1.9 were key ingredients of the Supreme Court’s 5–4 nuanced majority opinion in Perry v. United States. The court decided that the US government’s defaulting on its promise to pay Liberty Bond holders in gold coin was unconstitutional. But it also decided that the benign real returns on those bonds between 1933 and 1935 meant that bondholders had suffered no damages and therefore were entitled to no compensation. See Edwards (2018, 173–75).
42See the 1807 report of the Secretary of Treasury authored by Gallatin (1837) and Barro (1979) for a more formal treatment. For an analysis of the Barro tax smoothing model and its relationship to consumption smoothing models and other tax smoothing models, see
43Presidents Hoover and Roosevelt both pronounced rousing analogies between fighting wars and fighting depressions. Both presidents struggled with whether to invoke the 1917 wartime Trading with the Enemy Act for executive authority to take measures to fight the financial crisis and depression. Because during the last days of his administration he did not receive the political cover he sought from President-Elect Roosevelt, President Hoover declined to invoke the act in order to impose a bank holiday. President Roosevelt did invoke the act. See Edwards (2018, 27).
44The source of the weekly data displayed in Figures 1.13 and 1.14 is the table Resources and liabilities of each Federal Reserve Bank at the close of business on Fridays reported in each issue of the Federal Reserve Bulletin.The sharp increase in gold reserves on the assets side and Federal Reserve notes in circulation on the liabilities side of the balance sheet that occurred on June 22, 1917, is due to an amendment of the Federal Reserve Act, which permitted Federal Reserve Banks to count gold held by its Federal Reserve Agent as part of its required note reserve. Prior to this amendment, the liability of Federal Reserve Banks on outstanding Federal Reserve notes was reduced by the amount of gold held by the Federal Reserve Agent instead of the gold being considered as a collateral reserve.For the Federal Reserve’s assets, before the passage of the June 1917 amendment, gold reserves are the sum of gold coin and certificates in vault, gold settlement fund, and gold redemption fund. After this amendment, gold with Federal Agents and gold with foreign agencies are included in the gold reserves. Gold reserves include gold held against Federal Reserve notes starting in November 1919, and gold and certificates held by banks starting in December 1923. Reserves other than gold are legal tender notes, silver, and so on. Discounted US government securities are bills discounted: secured by US government obligations, and discounted private securities are bills discounted: all other. US government securities bought outright include bills bought in open market, US government bonds, US Victory notes, Treasury notes, and US certificates of indebtedness. All other assets are the sum of nonreserve cash, all other earning assets/other securities, bank premises, gold in transit or in custody in foreign countries, due from other Federal Reserve Banks (in transit), due from foreign banks, uncollected items, Federal Reserve notes, net assets, 5 percent redemption fund against Federal Reserve Bank notes, all other assets, and Federal Deposit Insurance Corporation stock.For the Federal Reserve’s liabilities, reserve deposits are the sum of government deposits/ US Treasurer general account, due to members—reserve account/member bank reserve account, due to nonmember banks clearing account, deferred availability items, reserved for government franchise tax, foreign bank, and other deposits. Prior to the June 1917 amendment, Federal Reserve notes are all notes and banknotes in circulation, net gold held by Federal Reserve Agents. After June 1917, Federal Reserve notes records all Federal Reserve notes and banknotes in circulation. All other liabilities are the sum of capital paid in/capital accounts, surplus funds, special deposits, reserve for contingencies, and all other liabilities.
45This section is intended to complement and supplement the account of Schuker (1988).
46The Treasury and the World War Foreign Debt Commission recognized these haircuts implicit in the renegotiated payment schedules. In Anne 1.3, we reproduce a table from the US World War Foreign Debt Commission (1927, 443) reporting the present values of the rescheduled repayments using constant discount rates by country. Depending on the discount rate used, the aggregate haircut, as measured by the Commission’s own calculations, ranged from one-fourth to one-half.
47We use Hicks-Arrow prices that Hall and others (forthcoming) inferred from market prices and quantities of Treasury bonds.
48The nominal value at date t is
where rs,s+1 is the nominal return on the portfolio between months s and s + 1. The real value at date t is
where πs,s+1 is the inflation rate between months s and s + 1. Thus, the real value is reported in December 1919 dollars. Prior to President Hoover’s debt moratorium in 1932, we computed the discounted value of the stream of promised payments; after the moratorium, we computed the perfect foresight discounted value of the stream of actual future payments.
49Joint Resolution of Congress, No. 133, approved by President Hoover on December 16, 1929.
50That is, the tax rates of 1.5 percent on the first $4,000 of taxable income, 3 percent on the next $4,000, and so on were reduced to 0.5 percent on the first $4,000 of taxable income, 2 percent on the next $4,000, and so on.
51See page 24 of the 1929 Annual Report of the Secretary of the Treasury on the State of the Finances.
52Ljungqvist and Sargent (2018, ch. 8) provide definitions and analysis of Arrow date-state prices.
53In the 11 years after the peak level of debt, the reduction had been only 25 percent.
54After the Civil War, Congress lengthened the average maturity of the debt. As chair of the World War Foreign Debt Commission, Secretary Mellon presided over an increase in the average maturity of the foreign debt owed to the US.
55There is an active theoretical literature on optimal maturity structures of government debt. Faraglia and others (2014), Bhandari and others (2017a,b), and Aguiar and others (2016) survey and contribute theories of the optimal maturity structure of government debt in settings that disarm the maturity-structure-is-irrelevant Modigliani-Miller theorems that prevail in complete market models. Faraglia and others (2014) and Bhandari and others (2017a,b) focus on (time-inconsistent) Ramsey plans in incomplete markets settings. Aguiar and others (2016) study Markov perfect plans in which debt dilution opportunities induce governments to issue only short-term debt along an equilibrium path.
57The calculations here are based on equation (1.1.6) of Anne 1.1 and are ex post in contrast to the ex ante calculations of Hilscher, Raviv, and Reis (2017).
58This means that we have subtracted from the total market values the values of securities held by Federal Reserve Banks and US government agencies.
59The terms in equation (1.1) leave a residual. This residual will include any mismeasurement of the government’s accounts, approximation errors in our accounting, and changes in the value the government’s other assets (for example, the gold stock, its vast land holdings, the nation’s railroads) that we omit in our analysis.
60Also see footnote 82 of Anne 1.1.
61By not adding risk premia to these rates to account for what were at various times doubtful prospects that these prospective amounts would be paid, we know that we overstate the value of these foreign claims. This fact affects the timing of the “foreign payouts” line in Figure 1.28 but not its main features and not its beginning and ending values.
62See footnote 84 in Anne 1.1 for an explanation of how we discounted promised payments beyond the horizon of our Hicks-Arrow price vectors.
63These assets are recorded in the Federal Reserve balance sheets reported each month in Board of Governors of the Federal Reserve System (1915–40).
65In 1932, representatives from Great Britain, Germany, and France met in Lausanne, Switzerland, to formulate a plan for reducing both the German reparation payments and the Allied war debts to the US. The US Senate rejected that debt reduction plan in December 1932, so the World War I foreign credits and the accumulated interest in arrears remain on the US Treasury’s books today.
66For descriptions and accounts of the German hyperinflation, see Sargent (1982) and Taylor (2013). See Schuker (1988) for a description of President Roosevelt’s ambivalent attitude about protecting the interests of American creditors in Germany.
67A concomitant effect was a huge redistribution from German domestic private creditors to German domestic private debtors.
68In April 1921, the Allies owed the US $9.4 billion in principal, with annual interest payments of $470 million.
69In 1919 and 1920, there had been substantial transfers to Germany by residents of the UK, US, and other countries speculating on a recovery of the German mark. John Maynard Keynes was among them. Some estimates of the funds transferred to Germany by the failures of those investments are as large as $3.5 billion. See Taylor (2013, ch. 14, Boom).
70See Table 2 in Klug (1993, 6).
72See Tooze (2014, 465).
73September 15, 1932. The New York American was a Hearst newspaper with masthead slogans “America First” and “An American paper for the American people.”
75Meanwhile, through a clever German government program designed by Reichsbank President Hjalmar Horace Greeley Schacht, German firms purchased roughly one-third of Germany’s dollar-denominated debts at heavily discounted prices between 1932 and 1940. These transactions are recorded in German Registration Office for Foreign Debt (1932–40) and discussed in Klug (1993) and Tooze (2006, ch. 3).
76After World War II, Germany resumed payments on these bonds. At the London Debt Conference of 1953, principals of the American tranches of the Dawes and Young loans were refinanced at 5.5 percent and 5.0 percent, respectively, with their maturity dates being extended to 1969 and 1980, respectively. Interest in arrears was refinanced into 20-year bonds paying 3 percent. See United Nations (1959, Annex I, 96). Holders of outstanding German commercial and municipal debts covered by the agreement had their claims written down to 38 percent of their 1931 level. See Klug (1993, 54) and Guinnane (2004).
77In the second half of 1933, Czechoslovakia, Great Britain, Greece, Italy, Latvia, and Lithuania made partial payments.With the exception of a postponement in 1941 and 1942, Finland continued to meet its scheduled obligations, even in 1943 and 1944 when it was allied with Germany. After 1949, the US Treasury dedicated Finland’s debt payments to finance educational exchanges between the US and Finland. Finland made its final payment in 1977.Greece made partial payments on its debt until 1942. Then in 1964 Greece agreed to an 80-year repayment schedule at 2 percent interest. Greece remains current on this loan. Its last installment is due on November 17, 2048.Hungary also made partial payments during the 1930s and early 1940s. In 1977, it paid the US Treasury over $4 million to become current on its obligations. Hungary repaid its debt in full in the early 1980s. The final payments were used to finance cultural exchanges between the two countries.Together the debts of Finland, Greece, and Hungary were less than four-tenths of 1 percent of the total indebtedness of foreign governments to the United States arising from World War I.
78When the government has issued perpetual consols, nt = ∞.
79Hall and others (forthcoming) computed nonlinear least squares coefficients α^i,t in the “level-slope-curvature” approximation qt+jtexp((α0,t+α1,tj+α2,tj2)j), then approximated qt+jtbyexp((α^0,t+α^1,tj+α^2,tj2)j).
80We also assume that qtjt=1forj1 for j ≥ 1 to express that a claim to a dollar does not expire at its maturity date.
81We assume that (qt+j1tqt+j1t1)=1 for j = 0 so that rtj,t0=0.
82In situations in which the payout stream is uncertain, a possible pricing theory is instead
where {s˜t+jt} is a promised payout stream and {π˜t+jt} is a sequence of fractions of promised payouts that a representative risk-neutral investor expects will actually be paid.
83To recognize that budget constraint (1.1.12) is equivalent with (1.1.6), use the definitions and approximation qt+jt1st+j1t1=Bt1j,Bt1=Σj=1nBt1jMt=stt,(νtνt1qt+j1tqt+j1t1Yt1Yt1)rt1,tjπt1,tgt1,t.
84Tables 1.5 and 1.7 of IMF (2014) report what the IMF calls gross financing needs of all member countries for the coming years. This is the sum of our first and third terms.
85When the horizon of the promised stream of foreign payments exceeds the horizon of our Hicks-Arrow prices (that is, when mt > nt ), we extrapolate our Hicks-Arrow prices by setting
where ρn,t is the yield to maturity of a nt -period risk-free zero coupon bond.

    Other Resources Citing This Publication