The previous chapter left the classical gold standard at its zenith in 1913: a fixed-rate monetary order spanning the industrial world, anchored in London, undergirded by central-bank cooperation and capital mobility. This chapter narrates its destruction. War broke its price-stability anchor in 1914; the postwar attempt to rebuild it ran from the German hyperinflation of 1922-1923 through Britain's 1925 return to pre-war parity and the gold-exchange standard of 1925-1929; the Depression of 1929-1933 collapsed the rebuilt structure; the 1930s became a policy laboratory in which three durable templates emerged from improvisation under crisis: American demand-management, German autarky, and Scandinavian countercyclical fiscal policy. By 1939 the monetary order of 1913 was unrecoverable. What replaced it is the next chapter's subject.
Named literature: Bresciani-Turroni (1937); Sargent (1982); Feldman (1993); Moggridge (1972); Eichengreen, Golden Fetters (1992); Eichengreen-Sachs (1985); Friedman & Schwartz, A Monetary History of the United States (1963); Keynes, Economic Consequences of the Peace (1919), Tract on Monetary Reform (1923), Economic Consequences of Mr. Churchill (1925), General Theory (1936, forward-pointer); Fisher, “Debt-Deflation Theory” (1933); Bernanke (1983, 2000); Galbraith, The Great Crash 1929 (1955); Skidelsky, Keynes biography; Cha (2003); Nakamura (1994); Maddison Project (Bolt & van Zanden); Mitchell, International Historical Statistics; Feinstein-Temin-Toniolo (1997); Wagenführ; League of Nations, World Production and Prices / Memorandum on Currency; Bairoch; Federal Reserve indices; ONS; Romer unemployment series; NBER.
On the morning of Friday, 31 July 1914, with Austria-Hungary mobilizing against Serbia and the European powers sliding toward general war, depositors queued outside the Bank of England in Threadneedle Street to convert notes into gold sovereigns. The run was orderly but accelerating. Over the weekend the Cabinet, the Treasury, and the Bank’s Court of Directors took the decision that ended the longest period of stable convertibility in modern history: an extended Bank Holiday, an emergency moratorium on debts, and the issue of Treasury notes that depositors could not exchange for gold. By the time the City reopened on the following Friday, the classical gold standard was suspended in Britain. Within days every other major belligerent followed.
The suspension was a deliberate policy act, not an accident of war. A working gold standard requires that any holder of paper claims on the central bank can present those claims and receive gold at a fixed price; this is the institutional commitment that makes the monetary order credible. War finance is incompatible with that commitment. Modern industrial war required spending on a scale (shells, ships, conscript pay, allied subsidies) that no belligerent could fund through taxation alone in the time available. The state share of GDP, which had stood at roughly 10 percent across the major European economies before 1914, climbed to between 30 and 50 percent at peak wartime mobilization. The British state in 1917 was spending more in a fortnight than it had spent in the entire fiscal year 1913. Britain’s wartime fiscal arrangements illustrate the general pattern: taxation expanded sharply under McKenna’s Treasury, but never to a level approaching wartime expenditure; the rest filled by debt issuance, much of it discounted at the Bank, and by direct Treasury borrowing. Taxation insufficient, debt monetized, money supply expanding, prices rising. A central bank obliged to convert paper into gold at pre-war parity could not finance such a war; the obligation had to give.
What followed was generalized inflation, in different magnitudes, across every belligerent. The price level by the war’s end stood between two and three times its 1913 level in most of the combatant economies, with the heaviest mobilizers and the weakest tax bases at the high end of the range.
| Country | Price level (1913 = 100) |
|---|---|
| United Kingdom | 203 |
| United States | 194 |
| Germany | 245 |
| France | 339 |
| Italy | 413 |
| Russia | 1100+ |
The British and American doublings reflected the discipline of large tax bases and conservative central banks; the French tripling reflected the cost of fighting on home soil with a smaller fiscal base; the German rise was held below the French through 1918 by price controls and rationing more than by monetary discipline; Italy’s quadrupling and Russia’s order-of-magnitude collapse marked the upper bound of what wartime monetization could do to a price level in four years.
Three institutional facts followed the price-level rises into peace. First, the suspension of convertibility was not formally revoked at the armistice; it persisted as the new normal, awaiting the policy decisions that would eventually re-establish (or not) the pre-war regime. Second, the wartime expansion of state capacity, tax administration, public-debt markets, central-bank operating practice, and statistical apparatus, did not contract back to pre-war levels; postwar state shares of GDP stabilized well above the 1913 baseline, a structural counter-thread to the chapter’s monetary-collapse main thread. Third, the price-level multipliers themselves left a problem for any postwar restoration: returning to gold at the pre-war parity would require either accepting the new price level (which would mean devaluing the currency in gold terms) or deflating the economy back to the 1913 price level (which would mean depressing wages, output, and employment for as long as the deflation took to grind through). That this latter path was attempted (by Britain at the pre-war parity in 1925, by France at a devalued parity in 1928, by Germany after the hyperinflation of 1923) is the puzzle the next two sections inherit.
In August 1922, a four-pound loaf of rye bread in Berlin cost about 70 paper marks. By August 1923 the same loaf cost roughly 2 million marks. By mid-November 1923 the bread cost between 200 and 400 billion marks, depending on the bakery, the hour of the day, and the most recent intra-day depreciation of the mark against any foreign currency. Workers were paid twice daily and ran from the pay window to the shops; restaurants quoted prices that increased between the moment of ordering and the moment of paying; a worker with a wheelbarrow of marks could lose the marks and keep the wheelbarrow and come out ahead. Between those two August figures, the price of bread had risen by a factor of roughly 5 billion. By November the cumulative collapse measured against the 1914 mark stood at the canonical figure: the November 1923 mark was worth approximately one trillion times less than its pre-war predecessor: a ratio of 1012, the textbook case of hyperinflation in the sense Phillip Cagan would later define (a price level rising more than 50 percent per month).
Behind the price chart sits an institutional story. The Reichsbank under Rudolf Havenstein, president from 1908 until his death on 20 November 1923, treated its statutory independence as license to discount Treasury bills on whatever scale the Treasury demanded, and the Treasury demanded a great deal. Germany’s tax base had been wrecked by the war and never reconstituted; the 1919 Versailles reparations settlement obligated Germany to substantial cash transfers in gold or hard currency to the Allied governments; the Ruhr crisis of January 1923—French and Belgian occupation of Germany’s coal-and-steel heartland in response to a missed reparations delivery—destroyed what remained of revenue from the country’s most productive region; the Reich responded by paying striking Ruhr workers and continuing reparations payments through fresh paper. The mechanism, presented at its sharpest, runs as follows. The Reichsbank monetized growing fiscal deficits by discounting government paper at the discount window. The expansion of the money supply produced inflation. The inflation produced expectation of further inflation. The expectation produced a sharp acceleration in the velocity of money (the rate at which a given currency unit changed hands), as holders of marks rushed to convert them into goods, foreign currency, or anything else before the next depreciation. Velocity acceleration is what carries a currency from steady inflation into hyperinflation; once expectations break, the same volume of paper will buy progressively less because nobody will hold it for any length of time.
The smaller cousins on Figure 12.1 (Austria, Hungary, Poland) confirm the pattern in different fiscal-political contexts. Austria, reduced by the postwar settlement to a rump state with a wholly inadequate revenue base, hyperinflated through 1922 with peak depreciation in August; the government accepted external supervision in October 1922 in exchange for a League of Nations stabilization loan, and the krone stabilized by year-end. Hungary followed a slower trajectory through 1922 and 1923 with peak depreciation in summer 1924; here too a League loan and external supervision (the Jeremiah Smith mission, on behalf of the League) underwrote the stabilization in March 1924. Poland, reconstituted from three former imperial territories with three different currencies and three different administrative traditions, hyperinflated through 1923 and stabilized in January 1924 under finance minister Władysław Grabski, who introduced a new currency (the zloty) backed by a balanced budget and a new central bank constitutionally barred from financing fiscal deficits. Each of the four cases stabilized when the same thing happened: the central bank’s discount window stopped accepting government paper, and the government’s fiscal accounts stopped relying on the discount window for their balance.
The lesson, given canonical statement by Thomas Sargent in “The Ends of Four Big Inflations” (1982), is that hyperinflations end when the fiscal regime changes, specifically when the central bank’s commitment not to monetize deficits becomes credible to the public holding the currency, and not when monetary aggregates alone change. The older textbook reading, in which the Reichsbank simply “printed too much money” and Schacht stopped printing, gets the chronology backwards. Hjalmar Schacht, appointed Reichsbank president in late November 1923 in succession to Havenstein, stabilized the mark by tying the Reichsbank’s hands and tying the Treasury’s as well. The November 1923 Rentenmark (a parallel currency notionally backed by a mortgage on German real estate, in practice backed by Schacht’s commitment that no further Rentenmarks would be issued for fiscal-deficit financing) and the simultaneous Reichsbank decision to refuse new discounts of Treasury paper, marked the end of the discount-monetization regime that had produced the inflation. The mark stabilized within weeks. The same fiscal-regime-change mechanism stabilized Austria in October 1922, Hungary in March 1924, and Poland in January 1924; the chronology in each case runs from credible fiscal commitment to monetary stability, not the reverse. Keynes’s Tract on Monetary Reform (1923), responding directly to the German hyperinflation experience, gave the broader policy critique of doctrinaire pre-war-parity restoration that the next section’s British case will recall, intellectual context worth following on the history of economic thought timeline.
Stabilization left bills to be paid politically. Middle-class savings denominated in marks, bank deposits, war bonds, life-insurance policies, mortgages held as assets rather than liabilities, had been wiped out, in many cases reduced to literally nothing, while debts denominated in marks were correspondingly liquidated. The German Mittelstand, the salaried professional and small-property-owning middle class whose savings the inflation destroyed and whose subsequent political alienation would matter very much in the late 1920s and early 1930s, had been pauperized in real terms while industrial debtors and large landowners had effectively defaulted on their mark-denominated obligations at zero cost. The Stresemann coalition that had held through the worst of the 1923 crisis fractured in November of that same year. The political costs of the German stabilization belong to a story the chapter will not narrate; what belongs here is the institutional fact that the stabilization was achieved through fiscal regime change, in the German case as in the smaller cousins, and that the Sargent reading of how this happened is the consensus among economic historians.
On 28 April 1925, in the Budget speech that opened his first year as Chancellor of the Exchequer, Winston Churchill announced that Britain would resume gold convertibility at the pre-war parity of $4.86 to the pound, effective immediately. The decision had been gestating in the Treasury and the Bank of England since 1918; it had been recommended by the Cunliffe Committee in 1919 and reaffirmed by the Chamberlain-Bradbury Committee in 1925; it had been pushed by Bank Governor Montagu Norman, who believed, as a matter of monetary statesmanship rather than as a matter of empirical economics, that London’s status as the world’s financial centre required restoration of the pre-war parity. Within weeks of the Budget speech, Keynes’s pamphlet The Economic Consequences of Mr. Churchill appeared, naming the consequences before they materialized: an overvalued sterling would force British prices and wages downward; coal, cotton, and shipbuilding, already weakened by structural shifts that pre-dated 1925, would face the brunt of the deflation; unemployment in the export industries would persist as long as the parity held. The empirical work of subsequent decades, particularly Donald Moggridge’s British Monetary Policy 1924-1931 (1972) and Barry Eichengreen’s Golden Fetters (1992), converged on an estimate of sterling overvaluation in 1925 of roughly 10 to 15 percent against the dollar at the announced parity.
The choice was political-symbolic rather than analytically grounded. The Treasury knew that British prices had not fallen back to 1913 levels; it knew that returning at $4.86 would force further deflation; it knew that the deflation would fall hardest on coal and the export trades. What the Treasury and the Bank wanted was the restoration of monetary normalcy as a fact, with London at the centre of it, a fact whose value they took to be self-evident even at the cost of a few years of high unemployment in the coalfields. Churchill, by his own later account, was not enthusiastic about the decision and was not in a position to override the Treasury and the Bank on a question they treated as technical. Keynes’s pamphlet, appearing in the Evening Standard in three instalments in July 1925 before its book publication, did the analytical work the Treasury had not done in public: a 10 percent overvaluation, he argued, would require a 10 percent deflation of British wages, which the British labour market would not deliver smoothly, which would produce unemployment, which would persist for as long as the parity was held. The analytical case was made in market time and lost in market time; the parity stood until September 1931.
The most visible domestic cost arrived in 1926. Coal-industry employers, facing the new parity’s squeeze on export prices, demanded wage reductions and longer hours; the miners refused; the dispute escalated through the Samuel Commission’s mediation attempts to a national general strike from 4 to 12 May 1926. The strike collapsed within nine days under Baldwin’s firm response, but the miners themselves stayed out for six months, returning to work on the employers’ terms. The general strike was the political symptom of an economic mechanism the Keynes pamphlet had named the previous year. Coal was the binding constraint because coal had been the export industry on which the British nineteenth-century trade surplus had been built; coal in 1925 was facing structural displacement (oil, electrification of marine propulsion, German and American competition in seaborne markets) on top of the new parity’s deflationary squeeze; coal, accordingly, was where the deflation bit first.
The British case sat inside a broader structural problem. The reconstructed gold standard of 1925-1929 was not the pre-1914 system rebuilt; it was a different architecture that the Genoa Conference of 1922 had recommended and that the major central banks had assembled piecemeal between 1922 and 1929. In the gold-exchange standard, central banks held reserves not only in gold but in claims on the major reserve currencies (sterling and the dollar), which were themselves convertible to gold. The arrangement economized on physical gold (a constraint that had become binding as world output grew faster than gold-mining output) by pyramiding the system on two reserve-currency centres. It also imported two new fragilities. First, the reserve-currency centres carried strain that pre-1914 London had carried alone, in a system where the Bank of England had been cooperating with two roughly equal partners (Reichsbank, Banque de France) rather than trying to coordinate with a Federal Reserve that did not yet know what kind of central bank it was and a Banque de France that under Moreau was actively accumulating gold. Second, the burden of adjustment fell asymmetrically on deficit countries. A country running a balance-of-payments deficit lost gold and had to deflate; a country running a surplus accumulated gold and was under no comparable pressure to inflate. France and the United States, by accumulating gold without allowing the corresponding domestic monetary expansion, sterilizing the gold inflow rather than letting it run through to prices and wages (contrary to the pre-1914 price-specie-flow mechanism Hume had described and that economics ch.16 walks formally), turned the international monetary system into one in which the deficit countries had to do all the adjusting and the surplus countries did none. The pyramided sterling balances, claims on London held by smaller central banks as reserve assets, were the channel through which any loss of confidence in London would feed back into the British gold reserve directly.
By 1929 the reconstructed system was in place and looked, on paper, much like its pre-1914 ancestor. Sterling was convertible at the pre-war parity; the dollar at $20.67 per ounce; the franc at the devalued 1928 Poincaré parity; the mark on the post-1924 Dawes Plan parity. Whether this architecture could survive a serious shock was a question with an answer the system was about to give.
On the evening of Sunday, 20 September 1931, the Bank of England suspended the Gold Standard Act, the second time in seventeen years that London had taken the decision that ended the world’s anchor currency’s convertibility. This time it was not war; it was a banking-and-payments crisis that had spread from a Vienna bank failure in May to Germany in July to the British reserves in August, and that had reached the threshold at which the Bank’s remaining gold could no longer underwrite the parity. By morning the decision was law. Within weeks, more than a dozen smaller economies had followed sterling off gold and pegged to it; within five years, all but the French gold bloc had abandoned the parity. The classical-and-reconstructed gold standard of 1815-1931 was over.
The chapter’s account works backward and forward from this date. Backward to the trigger: the New York stock-market crash of late October 1929 had been the proximate event, but it had been a US event first, with international consequences that took two years to develop. The Federal Reserve under Roy Young, then George Harrison at the New York Fed, had been tightening through 1928 and into 1929 in response to the speculative boom; the Dow lost roughly 23 percent in two days in late October and continued falling into 1932, ultimately bottoming roughly 89 percent below the 1929 peak. A stock-market crash by itself need not have produced a global Depression. What turned a US recession into the contraction of 1929-1933 was the conjunction of the crash with serial banking panics in the United States (four waves of panics between October 1930 and March 1933, in which roughly a third of all US banks failed and the money supply contracted by approximately 30 percent), with the international transmission of contraction through the gold-standard linkages that bound the major economies, and with the absence of any effective international or domestic lender of last resort prepared to break the chain.
The Depression’s depth was not uniform. Peak-to-trough industrial production fell by approximately 47 percent in the United States, 42 percent in Germany, 31 percent in France, 16 percent in the United Kingdom, and 8 percent in Japan. This cross-country variation, visible on Figure 12.2, is what the four-position debate in the next section is trying to explain. It is also visible spatially on the GDP map, where the 1914-1945 frame carries country-by-country GDP-per-capita trajectories with year-event annotations for the major economies around 1929 and 1933.
The 1931 sequence was the institutional turning point. The Creditanstalt, Vienna’s largest commercial bank and a relic of the pre-1914 Habsburg banking system, suspended payments on 11 May 1931 after the disclosure of large unrealized losses on its industrial-equity portfolio. The Austrian government and the Austrian National Bank attempted a rescue, then a foreign-loan rescue, then an emergency loan from the Bank for International Settlements; each succeeded in slowing the run rather than stopping it. Germany’s major banks were heavily exposed to Austrian counterparties and to short-term foreign deposits that were rapidly being withdrawn; through June and into July the German banking system experienced an accelerating outflow that climaxed in the Danatbank failure on 13 July 1931 and the Reichsbank’s suspension of foreign-exchange convertibility within a fortnight; exchange controls became the de-facto German exit from gold even though the official mark parity was never formally devalued. With Germany off gold in everything but name, the Bank of England’s reserves came under sustained pressure: holders of sterling balances (foreign central banks, foreign commercial banks, foreign deposit-holders in London) converted into gold at the announced rate, and the Bank’s reserves fell by approximately £200 million in the two weeks after 21 August. President Hoover’s standstill agreement on inter-Allied debts (announced 20 June 1931) removed one source of pressure but did not affect the immediate run on sterling. By the weekend of 19-20 September the MacDonald National Government, four weeks old and convened explicitly to defend the parity at the cost of public-spending cuts the previous Labour government would not deliver, took the decision to suspend convertibility. The narrative spine through these four months (Creditanstalt, Danatbank, sterling) is documented at length in Eichengreen’s Golden Fetters chapter 9, and it follows the canonical institutional shape: a peripheral bank failure exposes a core bank’s balance sheet, the core bank’s failure forces emergency measures by its central bank, the emergency measures spill across borders into a reserve-currency centre, the reserve-currency centre exhausts its gold defence, the parity goes.
By the morning of 21 September 1931 the global gold standard was broken. The remainder of the chapter is two questions: why was the Depression so deep (the next section’s historiographical surface), and what determined recovery (the section after that, then the policy-laboratory section, then the chapter’s verdict). For the broader intellectual context in which the four answers to the first question sit, the history of economic thought timeline carries the Great Depression as one of the era’s organizing events.
Why was the Depression so deep, and so global? The economic-history literature carries four major answers, each developed at length, each defended by serious scholarship, each making a different mechanism the load-bearing one. The chapter surfaces all four at strongest form in this section before taking a position in the chapter’s closing section. A reader who finishes this section knowing which reading the chapter favours has been given more by the prose than the prose intends.
The first reading is the Friedman-Schwartz monetarist account, given canonical statement in Milton Friedman and Anna Schwartz’s A Monetary History of the United States 1867-1960 (1963), and particularly in chapter 7, “The Great Contraction.” The contraction of the US monetary base by roughly a third between 1929 and 1933 was the proximate cause of the Depression’s depth in the United States, and the Federal Reserve’s failure to expand the base, to act as the lender of last resort that the Federal Reserve Act of 1913 had created the institution to be, turned a serious recession into a catastrophe. Specific failures: the Bank of United States in New York, a large commercial bank with a misleadingly federal-sounding name, was allowed to fail in December 1930 by a New York Fed that had the discretion and the resources to organize a rescue and chose not to; the spring 1931 opportunity for sustained monetary easing was missed; the response to sterling’s gold exit in September 1931 was a sharp tightening, by the standards of any monetary regime, that compounded the contraction at exactly the moment when the international transmission was already worsening. The Friedman-Schwartz reading at strongest form is not the claim that Federal Reserve passivity caused the initial 1929-1930 downturn; it is the claim that an active Federal Reserve, willing to extend the discount window and prevent the cascading bank failures of 1930–1933, would have prevented the recession from becoming a Depression. The 30-percent contraction of the US monetary base on Friedman and Schwartz’s reckoning is not a passive consequence of the panic; it is the policy variable on which everything else turns.
Friedman and Schwartz’s account is sharp on the Federal Reserve’s choices, but the second reading, Barry Eichengreen’s gold-standard account in Golden Fetters: The Gold Standard and the Great Depression, 1919-1939 (1992), asks whether the Federal Reserve had choices to make. The gold standard’s mentalité: the orthodox commitment to convertibility, internalized by central bankers and the political class as the foundation of monetary respectability. It bound central banks to contractionary policies precisely when expansion was needed. A central bank facing a domestic banking panic that wanted to expand its discount window would, in doing so, also expand its monetary liabilities and risk losing gold to other central banks; in the gold-standard logic, the appropriate response to gold loss was tightening, not easing. The contractionary policy of the Federal Reserve in 1931 and the parallel contractionary policies of the other major central banks were not analytical mistakes; they were institutional commitments under a regime that defined sound money as commitment to the parity. The international transmission of contraction, the Depression spreading from the United States through gold flows and trade balances to the rest of the gold-standard world, was the gold-standard system functioning as designed, propagating contraction symmetrically across all members. On Eichengreen’s reading, the Federal Reserve’s “failure” was not analytic mistake but institutional commitment; lifting the gold-standard constraint was the precondition for any expansionary policy. The recovery patterns documented in the next section (that countries off gold recovered, countries on gold did not) are the empirical signature of the binding institutional constraint at work.
Eichengreen’s gold-standard story explains the international transmission, but the third reading, the Keynesian-narrative account drawing on Keynes’s General Theory of Employment, Interest and Money (1936), Irving Fisher’s “Debt-Deflation Theory of Great Depressions” (1933), J.K. Galbraith’s The Great Crash 1929 (1955), and the historical-narrative tradition Robert Skidelsky’s Keynes biography stands within, asks what was being transmitted. The proximate fact of the Depression was the collapse of aggregate demand: total spending in the economy fell sharply between 1929 and 1933 and remained depressed through the 1930s. The mechanism was the collapse of investment under collapsed business expectations, what the General Theory would later name animal spirits (the irreducibly subjective component of investment decisions under genuine uncertainty), amplified by the contraction of consumption as wealth and income fell. Fisher’s distinct contribution, often underrated in subsequent surveys, was the debt-deflation mechanism: as prices fell, the real burden of nominal debts rose; debtors were forced into liquidation; liquidation forced further price declines; the spiral fed on itself, and only debt repudiation, debt reorganization, or a halt to deflation could arrest it. The Keynesian-narrative reading at strongest form is that monetary explanations require demand to fall; demand fell; the question of what made it fall and what kept it down requires the Keynes-Fisher analytical apparatus, not a substitute for it. (The intellectual development of the General Theory as a body of macroeconomic theory belongs to a different surface; the history-of-thought timeline carries it. The chapter cites Keynes here as a contributor to the historical-narrative tradition about the Depression itself.) Economics ch.16 walks the formal aggregate-demand framework that this paragraph compresses.
The Keynes-Fisher narrative captures the demand collapse, but the fourth reading is Ben Bernanke’s financial-fragility account, in “Non-Monetary Effects of the Financial Crisis in the Propagation of the Great Depression” (1983) and Essays on the Great Depression (2000), which asks why supply-side credit allocation broke too. The bank failures of 1930-1933 in the United States, and the parallel banking crises in central Europe, did more than contract the money supply (the Friedman-Schwartz mechanism) and depress aggregate demand (the Keynesian one). They destroyed irreplaceable information about who could be lent to. A bank’s knowledge of its borrowers (their cash flows, their character, their collateral, the local conditions in which their businesses operated) is built up over years of relationship and is mostly tacit. When a bank fails, that information is destroyed and cannot be reconstituted by a successor institution in any reasonable time. The cost of credit intermediation—the friction that separates the rate at which lenders are prepared to supply funds from the rate at which borrowers can effectively obtain them—rose sharply through the early 1930s, and stayed elevated long after the monetary contraction stopped. Firms with productive uses for capital found themselves unable to access it on terms reflecting their actual creditworthiness; new business formation collapsed; existing firms shed labour because they could not finance their working-capital needs. Bernanke’s reading at strongest form is not a refutation of the prior three but a fourth mechanism operating alongside them, particularly load-bearing for explaining the Depression’s length. Why recovery was so slow, even after monetary contraction stopped and aggregate demand began to revive, becomes legible only with the credit-channel mechanism in the picture.
Four readings, each making a different mechanism load-bearing: Federal Reserve passivity within US contraction (Friedman-Schwartz); gold-standard institutional commitment as binding constraint and international transmission belt (Eichengreen); aggregate demand collapse with debt-deflation amplification (Keynesian narrative + Fisher); credit-channel destruction explaining recovery’s slowness (Bernanke). The chapter’s closing section will take a position on which of these is doing the most work and where; the next two sections supply further empirical handles, the gold-exit-and-recovery pattern and the 1930s policy laboratory, that any verdict has to fit. BQ06 (central banks) takes the Friedman-Schwartz position as foundational for the modern monetary-policy debate; BQ08 (recessions) uses the four-reading framework as a template for thinking about subsequent contractions including 2008 and COVID. The intellectual landscape in which the four positions sit, Friedman in the counter-revolution, Keynes and Fisher in the Keynesian revolution, Eichengreen and Bernanke in the more recent modern-pluralism era cluster, is documented on the history-of-thought timeline.
Britain off gold in September 1931, recovery from 1932; the United States off gold in April 1933, recovery from mid-1933; France on gold until September 1936, no recovery until 1936. The pattern, given canonical statement in Barry Eichengreen and Jeffrey Sachs’s “Exchange Rates and Economic Recovery in the 1930s” (1985), is that recovery timing tracks gold-exit timing across the major economies. Figure 12.2’s vertical reference lines, color-matched to the country trajectories, make the relationship visible: the trough of each country’s industrial-production line sits within months of its gold-exit date, and the recovery slope after exit is roughly proportional to the time elapsed since exit. The empirical correlation is robust across alternative specifications (the result holds whether the recovery is dated from trough to a fixed reference year, from gold-exit to a fixed reference year, or measured as a continuous function of months since exit), and across alternative economy samples. Whether the correlation reflects gold-standard commitment as the binding institutional constraint (Eichengreen’s reading) or whether it reflects the freedom monetary authorities gained, after exit, to pursue the expansionary policies their domestic banking systems needed (the Friedman-Schwartz reading reformulated), is a question the chapter holds for the next section.
The mechanism on the gold-standard-commitment side runs as follows. A country tied to gold convertibility cannot expand its monetary base without losing gold to its trading partners; expansionary monetary policy and balance-of-payments pressure cannot coexist under fixed-rate convertibility, so a country that wants the first must take the second by abandoning the first. Once the parity is suspended, the central bank is free to discount on whatever scale the domestic banking system requires; the exchange rate adjusts (in 1931 sterling depreciated by roughly 30 percent against gold within months); the export industries gain immediate competitive advantage; the deflationary spiral that was forcing real-wage rises and debt liquidation is broken. The sterling area—the bloc of currencies that pegged to or shadowed the depreciated pound after September 1931, including most of the British Empire (excluding Canada), Scandinavia, Argentina, Portugal, and several others—reproduced this freedom across a substantial fraction of world trade.
The British case, taken on its own. The 21 September 1931 suspension was not accompanied by an immediately announced new parity; sterling floated, depreciated by approximately 30 percent against gold over the following months, and stabilized in mid-1932 around $3.40. The Bank rate, 6 percent on the day of suspension, was reduced through 1932 to 2 percent, a rate held until 1951 (with brief exceptions). The unemployment rate in the export industries fell sharply from 1932; aggregate UK industrial production was back above the 1929 level by 1934 and continued growing through the rest of the decade. Britain’s relatively shallow Depression (a 16 percent peak-to-trough fall in industrial production, against the United States’ 47 percent) reflected both the early exit and the more conservative pre-Depression position; the British boom of 1924-1929 had been muted, partly because the overvalued sterling that 12.3 narrated had constrained it, so the contraction had less to contract from.
The American case ran on a different schedule. From the second Hoover administration through the Roosevelt inauguration on 4 March 1933, the US gold standard was held under increasing strain; banking panics worsened through February 1933, by which point a substantial fraction of the country’s banks had closed at state-government direction. Roosevelt’s first official act, on 6 March 1933, was a national bank holiday closing all US banks for inspection and recapitalization; on 5 April the Executive Order 6102 prohibited private hoarding of gold and required US persons to deliver gold to the Federal Reserve at the official $20.67 rate; through the spring and summer of 1933 the dollar was effectively floating against gold; in January 1934 the Gold Reserve Act fixed the new official parity at $35 per fine ounce, a 41 percent devaluation against the pre-1933 parity. Industrial production began rising from mid-1933, doubled by 1937, suffered a sharp setback in the 1937-1938 “Roosevelt recession,” and recovered again through the rearmament boom of 1938-1939. The American recovery was real but incomplete: industrial production reached 1929 levels only briefly in 1937 before falling back, and unemployment remained above 14 percent at the close of the decade.
The French case demonstrates the mechanism in reverse. France, with Belgium, the Netherlands, Switzerland, Italy (until its 1935 Ethiopia adventure), and Poland, formed the gold bloc: the holdouts who maintained the gold parity through the early 1930s on the principle that monetary discipline was the foundation of financial credibility. The bloc’s economies suffered the Depression’s deepest sustained deflation, with prices falling continuously through 1935 and industrial production stagnating well below the 1929 level through the entire period. The Tripartite Agreement of September 1936—a coordinated devaluation by France, Britain, and the United States, brokered to allow France to exit gold without provoking competitive devaluation—marked the end of the gold bloc. The Popular Front government’s domestic monetary easing followed the franc’s 1936 devaluation, and French industrial production began rising for the first time in five years. Recovery, when it came, came on the same schedule the rest of the chapter has shown: from gold exit, not before.
Two corollaries follow from the gold-exit-recovery pattern. First, the Depression’s depth was not uniform: countries that left gold earlier suffered shallower contractions, countries with stronger lender-of-last-resort capacity (the Federal Reserve, when it eventually used its powers; the Bank of England, in the period after 1931) recovered faster, countries with greater banking-system fragility (the United States with its thousands of small unit banks; central Europe with its universal-bank exposure to industrial equity) suffered deeper banking crises and longer credit-channel disruption. Second, the recovery patterns documented here put empirical weight behind Eichengreen’s gold-standard reading, the differential timing is exactly what his account predicts, while leaving unresolved the question of whether the binding constraint was the gold standard as institutional commitment or the underlying monetary policy that the gold standard ruled out. The chapter holds that question for the verdict section. What countries did after leaving gold is the next section’s subject. For the further country-by-country detail this section’s 6-economy comparison cannot accommodate, the GDP map carries trajectories for roughly 180 economies across the same period.
The 1930s were the period in which activist macroeconomic policy was tried on a serious scale for the first time. Three durable templates (plus a fourth case that prefigured one of them outside the West) emerged from the experiments. A decade earlier, none would have been thinkable as peacetime policy; a decade later, all were embedded, in modified form, in the postwar consensus. This section walks the templates as a comparative survey, not as positions to choose between.
The US New Deal ran from Roosevelt’s March 1933 inauguration through 1938, with the second New Deal of 1935 shifting the programme from immediate-crisis response to longer-term institutional reconstruction. The financial-regulation half rebuilt the US banking and securities system on a framework that endured into the 1980s. The Banking Act of 1933, widely known as Glass-Steagall, separated commercial from investment banking and established the Federal Deposit Insurance Corporation, under which depositor losses in failed banks were federally guaranteed up to a limit. The Securities Acts of 1933 and 1934 imposed disclosure requirements on public securities and established the Securities and Exchange Commission. The Banking Act of 1935 reorganized the Federal Reserve under Marriner Eccles, centralizing monetary-policy authority in the Board of Governors. The demand-management half ran through the work-relief programmes (the PWA, CWA, and WPA) and the Agricultural Adjustment Acts that supported farm prices through acreage restrictions and parity payments. The National Industrial Recovery Act of 1933 was struck down by the Supreme Court in 1935; its labour provisions were re-enacted as the National Labor Relations Act, founding the modern American framework for collective bargaining. The Social Security Act of 1935 established the federal old-age insurance system that remains the largest single component of the US federal budget. Recovery through 1933-1937 was real but reversed sharply by the “Roosevelt recession” of 1937-1938, triggered by Treasury sterilization of gold inflows and simultaneous fiscal tightening; the lesson, absorbed into the postwar Keynesian consensus, was that countercyclical policy could not be wound down prematurely without restarting the contraction.
Schachtian autarky and rearmament in Germany ran on a different logic. Hjalmar Schacht, returned to the Reichsbank presidency in 1933 and serving simultaneously as Economics Minister from 1934 to 1937, designed the architecture by which Germany rearmed without provoking the foreign-exchange crisis that conventional financing would have produced. The New Plan of September 1934 imposed comprehensive exchange controls and replaced multilateral trade with bilateral clearing arrangements: country-by-country agreements in which German imports were paid for in offsetting German exports rather than in convertible currency. The arrangements concentrated on Latin American suppliers (Brazilian coffee, Argentine grain) and Balkan trading partners (Romanian oil, Yugoslav minerals) where Germany had bargaining leverage as a large counterparty. Internal financing ran through Mefo bills: off-balance-sheet promissory notes issued by a state-controlled shell company (Metallurgische Forschungsgesellschaft) to armament suppliers, discounted at the Reichsbank, repaid out of future bill issuance. The arrangement kept armament expenditures off both the Reich’s published budget and the Reichsbank’s published monetary aggregates. Wage and price controls held inflation suppressed; capital controls trapped private wealth domestically. In narrow economic terms the result was rapid recovery: industrial production reached 1929 levels by 1934 and roughly full employment by 1936, the fastest recovery among the major economies. The cost was the deepening structural commitment to rearmament and the bilateral trade fragmentation that closed Germany off from the multilateral system. The regime worked at what it set out to do, and what it set out to do was war.
Sweden under Stockholm school influence ran the third template. The intellectual cluster (Gunnar Myrdal, Erik Lindahl, Bertil Ohlin, Erik Lundberg) had been working through the 1920s on the theoretical questions (savings-investment imbalance, ex ante versus ex post categories, dynamic sequence analysis) that the Anglo-American economics profession would only catch up to after the General Theory. Per Albin Hansson’s Social Democratic government, taking office in September 1932, deployed the framework in policy. Ernst Wigforss, finance minister from 1932 to 1949, framed the 1933 budget around explicit countercyclical fiscal expansion (public works, agricultural support, unemployment relief), financed through deficit borrowing on the principle that the deficit was a stabilization instrument, not an emergency expedient. Sweden’s recovery was strong: industrial production was back above 1929 levels by 1934 and rose through the rest of the decade, with unemployment falling roughly in step. The Stockholm-school template, more than the New Deal’s improvisations or Schachtian autarky’s closed system, supplied the postwar Scandinavian welfare-state architecture and the doctrinal precedent for the postwar Keynesian fiscal orthodoxy that the Anglophone economies would adopt under the General Theory’s influence. The history-of-thought timeline carries the broader Keynesian-revolution context.
Japan’s case prefigured the early-exit template outside the West. Takahashi Korekiyo, appointed Finance Minister in November 1931, took Japan off gold on 13 December 1931, three months after sterling. Through 1932-1936 Takahashi ran the cleanest case of countercyclical policy in the period: substantial fiscal expansion (military procurement and rural-relief spending), accommodative Bank of Japan policy (direct underwriting of government bonds, on-sold to the public as inflation expectations recovered), and a deliberately undervalued yen supporting export-led recovery. Japanese industrial production was above the 1929 level by 1933, the steepest early-recovery line on Figure 12.2, and continued rising through the decade. Takahashi was assassinated on 26 February 1936 by junior army officers in the 2-26 coup attempt; his death removed the institutional check on military spending, and the Japanese trajectory passed under the control of the institutions that would carry it through the China war and the Pacific war. The clean policy-laboratory case ended with Takahashi.
The four templates leave inheritances that the next two chapters build on. Bretton Woods was designed by participants—Harry Dexter White, Keynes himself—whose policy formation lay in the New Deal and the British 1931-1939 reflation; the postwar mixed-economy consensus that chapter 14 documents has its institutional roots in the 1930s experiments; the Schachtian model’s clearing architecture survived into the postwar communist trade system without ever again being attempted in the West. Chapter 13 picks up the design of the postwar replacement.
The four readings of the Great Depression are not just historiography. They shape how the modern economics profession reads 2008, COVID, and every recession in between; how central banks calibrate their lender-of-last-resort interventions; how international institutions diagnose currency crises; how historians of capitalism understand the relationship between financial fragility and macroeconomic stabilization. A position on which mechanism was load-bearing in the Depression is, in practice, a position on what kind of economic policy is supposed to do under what kind of stress. The chapter takes a position. The position is conditional on the empirical material the previous sections surfaced; that conditioning is the qualifier, not throat-clearing.
Eichengreen’s gold-standard reading is load-bearing for the Depression’s international transmission and for the differential recovery patterns documented in 12.6. The cross-country variation in Depression depth and recovery timing tracks gold-standard exit timing too tightly to be coincidence; the institutional commitment to convertibility was the binding constraint that turned a US contraction into a global Depression and that determined which economies recovered when. This is the strongest single causal claim in the literature, and the recovery-timing evidence puts it on a footing that the alternative readings have to accommodate rather than dismiss.
Friedman and Schwartz’s monetary-contraction reading is load-bearing within the United States. The Federal Reserve’s choices between 1930 and 1933 are not exonerated by recognizing the gold-standard constraint, because the gold-standard constraint did not require the specific failures (the Bank of United States in December 1930, the spring 1931 missed easing, the autumn 1931 disastrous tightening) that Friedman and Schwartz document. The Federal Reserve had room within the constraint to act differently, and at decisive moments did not. The 30-percent contraction of the US monetary base is not predicted by the gold-standard story alone; it is the additional fact that explains why the US Depression was 47 percent deep against the British 16 percent. Within the United States, on the chapter’s reading, the Friedman-Schwartz mechanism is the load-bearing one.
The Keynesian-narrative reading is correct on the proximate fact. Aggregate demand collapsed and stayed collapsed across the major economies; monetary explanations of why this happened require something to explain how a monetary contraction translated into a sustained shortfall in spending; the Keynes-Fisher analytical apparatus supplies the translation. Within the Keynesian narrative, Fisher’s debt-deflation contribution deserves more weight than it usually gets: as a separately-named amplifier through which falling prices fed liquidation fed further falling prices, it carries explanatory load that pure aggregate-demand stories do not. The chapter restores Fisher’s standing here. (The intellectual development of Keynesian macroeconomics as a body of theory is a separate subject; the textbook’s economics volume carries it.)
Bernanke’s financial-fragility reading is load-bearing for Depression length. The credit-channel mechanism explains why recovery was so slow and so incomplete even after the monetary contraction stopped and aggregate demand began to recover: bank failures destroyed tacit borrower information, the cost of credit intermediation rose and stayed high, and productive firms could not access funding on terms reflecting their actual creditworthiness. The Bernanke account is not, on the chapter’s reading, a refutation of any of the prior three; it is a fourth mechanism operating alongside them, particularly important for explaining why the Depression’s tail extended through the second half of the 1930s.
The chapter’s position, then: the four readings are best understood as complementary, not rival, with each load-bearing at a different point in the causal chain. Eichengreen’s gold-standard story carries the international transmission and the cross-country recovery pattern. Friedman and Schwartz carry the US contraction depth. The Keynesian narrative, with Fisher restored to his proper weight, carries the proximate aggregate-demand collapse and the deflation-amplification mechanism. Bernanke carries the length and incompleteness of recovery. The earlier debates’ framing, in which Friedman-Schwartz refuted Keynes, Eichengreen refuted Friedman-Schwartz, Bernanke refuted Eichengreen, misreads the structure of the disagreement; each new account took its predecessors’ mechanisms as inputs to a more complete picture rather than as targets to be dismissed. The synthesis position is the chapter’s contribution to the historiography.
Three durable inheritances pass from the 1930s into 1939, and from 1939 into the order the next chapter narrates. Gold-standard orthodoxy is discredited as the default monetary architecture of the industrial world; no major economy will return to it after 1945, and the Bretton Woods system that replaces it builds in mechanisms the gold standard lacked. Activist macroeconomic policy, including countercyclical fiscal expansion, lender-of-last-resort central banking, deposit insurance, and securities regulation, is validated by the 1930s experiments and embedded in the postwar policy consensus across the Western mixed economies; the New Deal, the Stockholm school, and the British 1931-1939 reflation are the precedents the postwar settlement codifies. Regional currency-bloc geopolitics—the sterling area, the dollar bloc, the Schachtian clearing zone, the gold bloc—replaces the single integrated monetary order of pre-1914 with a fragmented monetary geography that the Bretton Woods designers will spend the 1940s trying to re-integrate. Each of the three inheritances structures the next chapter’s subject. Chapter 13 picks up what was built on them. BQ06 and BQ08 carry the chapter’s verdict into the modern central-banking and recession-comparison debates.