Chapitre 13 Théorie de la croissance

Intro

The chapter that ended at the Federal Reserve's 1914 founding ran through one war, the interwar collapse, and a second war that closed in 1945 with the United States holding roughly two-thirds of the world's monetary gold and the only large industrial economy left intact. From that position the postwar international economic order was designed. Bretton Woods, the Marshall Plan, and the General Agreement on Tariffs and Trade are the three institutional pieces this chapter walks. The chapter treats them as a coherent architectural system: each piece complemented the others, the design responded to specific lessons drawn from the interwar period, and the order operated for a quarter-century before the structural pressure that the gold-dollar standard had carried from the start brought it down. The chapter ends in 1971 at the cusp, with the architecture intact and the pressures visible. The collapse is the next monetary chapter's subject.

Named literature: Harrison (1998) The Economics of World War II; Milward (1977) War, Economy and Society 1939–1945; Boughton (multiple IMF historical works on Bretton Woods institutional history); Helleiner (1994) States and the Reemergence of Global Finance; Skidelsky (Keynes biography vol. 3); Steil (2013) The Battle of Bretton Woods; Eichengreen Globalizing Capital; James (1996) International Monetary Cooperation Since Bretton Woods; Hogan (1987) The Marshall Plan: America, Britain, and the Reconstruction of Western Europe; DeLong and Eichengreen (1993) “The Marshall Plan: history's most successful structural adjustment program”; Eichengreen and Uzan (1992); Bagwell and Staiger (2002) The Economics of the World Trading System; Irwin (US trade-policy history); Triffin (1960) Gold and the Dollar Crisis; Ruggie (1982) on embedded liberalism; primary documents (IMF Articles of Agreement; IBRD Articles of Agreement; GATT 1947 text; Atlantic Charter 1941; Beveridge Report 1942).

13.1 Le modèle de Ramsey-Cass-Koopmans

In 1939 the United States built roughly 6,000 military aircraft. In 1944 it built roughly 96,000. The same five years saw US output of tanks rise from a few hundred to over 17,000 a year, naval combat tonnage grow into a fleet larger than every other navy on earth combined, and the fraction of GDP devoted to war production climb from a peacetime trace to roughly 40 percent. This was not a marginal expansion. The American war economy of 1944 was a different industrial system from the American economy of 1939, built in less than five years on top of the surviving capacity of the depression decade.

The transformation ran on three financial instruments and one administrative one. The Revenue Act of 1942 expanded the federal income tax from a levy on a few million high earners to a mass tax on wage and salary income; payroll withholding, introduced in 1943, made the new tax administratively feasible. War bonds, sold in seven national drives between 1942 and 1945, absorbed roughly $186 billion in retail and institutional savings and turned a sizable fraction of household financial accumulation into Treasury obligations. Deficit financing (net Treasury borrowing) covered the residual; the federal debt rose from about 40 percent of GDP in 1941 to roughly 120 percent in 1945. The administrative instrument was the War Production Board (WPB), established January 1942, which directed the conversion of consumer industry to war production, allocated scarce raw materials (steel, aluminum, rubber, copper) by priority order, and issued the Controlled Materials Plan distributing inputs across thousands of prime contractors and subcontractors. The Office of Price Administration set ceiling prices on consumer goods and rationed sugar, gasoline, tires, meat, and shoes. Detroit's auto plants stopped producing cars in February 1942 and converted to tanks, aircraft engines, and military trucks; Henry Ford's Willow Run plant outside Ypsilanti was producing one B-24 Liberator bomber per hour by 1944.

The other belligerents mobilized within different constraints. Britain entered the war as a still-substantial industrial economy and exited as a financially exhausted one. Lend-Lease, the program enacted in March 1941 by which the United States supplied military and economic aid to Allied governments without immediate payment, transferred roughly $31 billion in goods and services to Britain by war's end (and roughly $11 billion to the USSR). The flow kept British war production financed but left Britain at war's end with depleted reserves and a negotiating position structurally weaker than its industrial output suggested. The USSR's mobilization ran on a different logic: in the late summer and fall of 1941, as German armies advanced toward Moscow, Soviet planners dismantled and shipped roughly 1,500 industrial enterprises east of the Urals, reassembling them in the Volga and Siberian regions over the following year. Soviet T-34 tank production, once the relocated plants reached operating scale, ran at roughly 14,000 to 15,000 units per year through 1943–1944, more than the combined Panther and Tiger output of German industry. (Soviet wartime production data has known uncertainty; Mark Harrison's reconstruction is the standard source, drawing on archival material that became available only after 1991, and the figures should be read with that caveat.) Germany under Albert Speer's armaments ministry achieved its peak output in mid-1944, the Speer-rationalization paradox in which production rose even as Allied bombing destroyed substantial capacity, by concentrating remaining resources on a narrowed set of weapons types. The 1944 German figures are gross production at the factory gate, not net delivered output. Japan mobilized within tighter resource constraints than any other major combatant; US submarine warfare destroyed the merchant fleet supplying Japanese industry with raw materials from the conquered Southeast Asian periphery faster than Japanese shipyards could replace it, and by 1944 the war economy was running down its raw-material stocks rather than replenishing them.

The table makes the structural point. By 1944 the United States alone was building more aircraft than Germany, Japan, and Italy combined, and US naval construction was an order of magnitude larger than any single Axis state's. Add British, Soviet, and Commonwealth output and the Allied-to-Axis ratio in finished munitions ran at roughly three-to-one across the headline categories. The war was decided on many fronts; the production ratios are the necessary background condition that made the decisions possible. Strategy, leadership, terrain, and the operational performance of armies in the field determined where and how the war was won. Industrial output determined that the war could be won at all on the timetable the Allies fought it.

The wartime experience shaped the postwar political consensus that ran through the 1940s and 1950s in the major Western economies. The successful state-led mobilization (the WPB's industrial coordination, OPA price controls, the Revenue Act of 1942's mass income-tax base, the deficit-financing-plus-war-bonds combination, the wartime conversion of consumer industry under government direction) demonstrated that large-scale state coordination of the economy could be administratively competent and politically tolerated when the goal was sufficiently agreed. The political class that emerged from the war was prepared to accept government roles in employment policy, social insurance, industrial planning, and macroeconomic stabilization that the political class of the 1920s would have rejected. The Beveridge Report, published in Britain in December 1942 under the title Social Insurance and Allied Services, sketched a comprehensive welfare-state architecture covering family allowances, a national health service, full-employment commitment, and contributory social insurance, all of which the Attlee government would implement after 1945. The Beveridge plan was not a postwar invention; it was wartime policy work, drafted while the bombs were still falling, and it carried the wartime legitimacy of state-coordinated provision into the peacetime political settlement. The mixed-economy consensus of the postwar quarter-century, with its Keynesian demand management, expanded welfare states, and industrial-policy commitments, rested on the wartime experience that made it credible. The institutional architecture that emerged from the war was designed by people who had just lived through both the production-coordination success and the interwar collapse it had to repair.

13.2 The Road to Bretton Woods: Interwar Lessons and the Anglo-American Negotiation

The architects of the postwar order had two reference points: a war they were still fighting and an interwar collapse they had just lived through. The first reference point set the political conditions for ambitious institutional design. The second specified what the design had to prevent.

The interwar lessons compressed into a short list. Competitive devaluation in the 1930s, with sterling off gold in 1931, the dollar devalued against gold in 1933–1934, and the Tripartite Agreement of 1936, had shown that uncoordinated exchange-rate adjustment turned every country's defense of its own balance of payments into a deflationary export of unemployment to its trading partners. Capital flight had destabilized the late-1920s gold-exchange standard and accelerated bank failures during the Depression's worst years. Bilateral clearing arrangements, particularly Hjalmar Schacht's New Plan of 1934, had carved trade into politically managed corridors and reduced the multilateral character of the international economy that the pre-1914 gold standard had supported. The gold standard itself, restored across most of the industrial world by 1928 and broken across most of it by 1936, had transmitted deflationary pressure from countries trying to defend parities to those that had already abandoned them, forcing competitive contraction across the system. (The interwar story is the previous chapter's subject.) The postwar architects took these failures as design constraints. Whatever the new order built, it had to prevent competitive devaluation, manage capital movements without forcing autarky, restore multilateral trade, and decouple domestic monetary policy from a metallic discipline that had collapsed under political pressure.

The Anglo-American negotiation that produced the design ran from 1941 through 1944 in two parallel channels. The political channel was opened by the Atlantic Charter, signed by Roosevelt and Churchill aboard the USS Augusta in August 1941, four months before the United States entered the war; its eight points included a commitment to “the fullest collaboration between all nations in the economic field” and to access “on equal terms” to trade and raw materials. Article VII of the Mutual Aid Agreement, signed in February 1942, made the Lend-Lease aid that was keeping Britain in the war conditional on a postwar British commitment to negotiate the elimination of trade discrimination, specifically of the imperial preference system established at Ottawa in 1932. Article VII pulled British Treasury planning into a postwar framework that the United States would substantially set the terms of. The technical channel was opened by parallel drafting work at the British Treasury under John Maynard Keynes and at the US Treasury under Harry Dexter White, which produced the two competing institutional plans that the conference would have to reconcile.

Keynes's plan was for an International Clearing Union. Its central instrument was the bancor, a true international reserve asset, defined in terms of gold but not redeemable in it, in which all international trade balances would be denominated and settled. Each country would have an overdraft facility at the Clearing Union sized to its trade volume; surplus countries would accumulate bancor balances and deficit countries would draw them down. Adjustment obligations were to be symmetric: persistent surplus countries would face penalties (interest charges on excess balances, ultimately confiscation of accumulated surpluses) parallel to those facing deficit countries, on the theory that the burden of adjustment in any imbalanced trading relationship belongs as much to the country running the surplus as to the country running the deficit. Capital controls were an integral part of the plan, treated not as an emergency tool but as a standing instrument that countries would deploy to insulate domestic monetary policy from speculative capital flows. The plan's premises were Keynesian in the institutional sense: full-employment commitment as a domestic policy priority that international arrangements should support rather than constrain; skepticism that free capital movement was compatible with the macroeconomic policy autonomy national governments would need; symmetric-adjustment intuition that placed responsibility on the system rather than on the weaker partner.

White's plan was for a Stabilization Fund. Its central instrument was a pool of national-currency contributions, denominated and effectively anchored in dollars, against which member countries could draw in conditional tranches when they needed financing for a balance-of-payments deficit. Adjustment was structurally asymmetric: deficit countries that drew on the Fund accepted policy conditions; surplus countries (for the United States in 1944, this meant the United States itself) faced no parallel obligation. The dollar was anchored to gold at $35 per ounce, a parity inherited from the 1934 devaluation, and other currencies were anchored to the dollar through declared par values. Capital controls were permitted but not affirmatively endorsed. The plan reflected the structural position of the United States as the war's dominant creditor: gold-rich, current-account surplus, the supplier rather than the user of international liquidity, the country that would write the rules and bear the lowest cost of having them enforced.

The political economy of the negotiation ran inside that asymmetry. Keynes negotiated from a position of bankruptcy: Britain by 1944 had liquidated most of its dollar reserves and accumulated sterling-area liabilities (the “sterling balances” held by India, Egypt, and other suppliers of wartime goods to the British war effort) that exceeded its remaining gold and dollar holdings by a wide margin. White negotiated from a position of dominance, with Treasury Secretary Henry Morgenthau Jr.'s political authorization and Roosevelt's broad policy framing behind him. The settlement reached at the Mount Washington Hotel in Bretton Woods, New Hampshire, in July 1944 was White's plan as the institutional skeleton. The bancor was rejected; the dollar-anchored par-value system became the architecture. Keynesian concessions were inscribed at specific points: Article XIV permitted a transition period (initially expected to last about five years, extending in practice to December 1958 for major Western European currencies) during which countries could maintain exchange restrictions; Article VI §3 affirmatively permitted member countries to use capital controls to regulate international capital movements, an explicit departure from the gold-standard premise that capital should move freely.

The conference itself ran for three weeks beginning July 1, 1944. Seven hundred and thirty delegates from 44 nations gathered at the Mount Washington Hotel; Keynes chaired Commission II (the Bank), White ran Commission I (the Fund). Most of the technical drafting had been done in advance through the bilateral Anglo-American channel; the conference's plenary work was substantially the formalization of pre-negotiated text and the management of the smaller delegations whose acceptance was needed for institutional legitimacy. The Final Act was signed July 22. (The intellectual-history timeline locates Bretton Woods within the larger arc that includes Keynes's published work and the keynesian revolution of the 1930s; for that placement see the keynesian-era cluster on the intellectual-history timeline. The doctrinal lineage from the General Theory through the postwar policy mainstream is C-book territory.)

The historiography of the negotiation contains an alternative reading worth naming. Benn Steil's The Battle of Bretton Woods (2013) presents the negotiation as a near-zero-sum US-versus-UK contest in which White outmaneuvered Keynes on essentially every contested provision, drawing additional weight from the documented thread of White's pro-Soviet sympathies and his later identification by Soviet defectors as a wartime intelligence source. The chapter's house line follows a broader literature (James Boughton's IMF institutional histories, Eric Helleiner's political-economy framing of the postwar financial order, Robert Skidelsky's three-volume Keynes biography) that reads the negotiation as a bilateral compromise constrained by structural facts the negotiators could not change: US gold dominance and creditor position, British wartime financial dependence, the political requirement that whatever was built had to pass the US Senate and a British Parliament that had just spent five years at war. Steil's account adds detail to the bilateral contest; it does not displace the structural reading on which the chapter's narrative rests.

What the settlement built was a coherent institutional architecture: three institutions with mutually-reinforcing mandates, designed to operate as a system.

13.3 Le modèle AK

The Bretton Woods conference produced not three independent institutions but one system designed to operate as a coordinated whole. The International Monetary Fund managed the par-value exchange-rate system and supplied emergency balance-of-payments financing. The International Bank for Reconstruction and Development supplied long-term capital for reconstruction and (later) development. The General Agreement on Tariffs and Trade, signed in 1947 as a provisional first step toward an International Trade Organization that never came into being, became the trade-rules regime by default. The three together carried the four functional commitments that the interwar lessons had specified: managed exchange-rate adjustment without competitive devaluation; balance-of-payments financing without forced deflation; long-term reconstruction finance; and progressive multilateral trade liberalization through reciprocal bargaining.

The IMF as designed sat at the system's center. Each member country declared a par value for its currency, expressed either directly in gold or in US dollars at the $35-per-ounce parity. Market exchange rates were maintained within ±1 percent of par value through central-bank intervention; the IMF supplied dollar liquidity for that intervention through drawing rights against the Fund's pool of subscribed currencies. Quotas, the financial subscriptions that determined each member's contribution to the Fund's pooled resources, its drawing-right entitlement, and its voting weight on Fund decisions, were calculated from a formula combining national income, gold and dollar reserves, exports, and trade variability. The initial quota distribution gave the United States roughly 32 percent of total subscriptions and roughly 27 percent of voting power; the United Kingdom held roughly 14 percent. The structural asymmetry was built into the institution: the United States held an effective veto over the qualified-majority decisions (changes to quotas, par-value adjustments above defined thresholds) that required 80 or 85 percent supermajorities, and US Treasury preferences carried disproportionate weight on every operational question. Fixed-but-adjustable exchange rates were the system's defining technical feature: par values were not permanent but could be adjusted by member countries with IMF consent in cases of fundamental disequilibrium, an undefined term whose lack of precise specification was deliberate, leaving the Fund and member countries discretion to negotiate adjustments rather than committing to a mechanical trigger. (The formal mechanics of fixed-but-adjustable exchange rates, including the Mundell-Fleming model and the open-economy Trilemma framework that organizes the policy trade-offs among capital mobility, exchange-rate stability, and monetary-policy autonomy, live in economics ch. 17; this chapter narrates the institutional operation. The monetary-theory background sits in economics ch. 16.) Current-account convertibility, defined as freedom to exchange currencies for the financing of trade in goods and services, was the operational goal, codified in IMF Article VIII. Embedded liberalism, the term Ruggie (1982) gave to the political compromise at the system's heart in which open international markets for goods and current-account transactions were paired with permitted closure of national capital accounts so that domestic full-employment policy retained autonomy, was inscribed in Article VI §3, which affirmatively permitted member countries to maintain or impose capital controls.

The IBRD was designed as a long-term lender for reconstruction and development. Its initial focus was European reconstruction; its first loans, made in 1947, went to France ($250 million, the largest single loan in its first decade), the Netherlands, Denmark, and Luxembourg. The Bank's lending terms were long-maturity at near-market interest rates, financed initially by paid-in capital subscriptions from member governments and then increasingly by bond issuance in private capital markets, particularly New York. As Marshall Plan grants substantially displaced reconstruction-finance demand from the major Western European economies after 1948, the Bank's portfolio shifted toward development lending in the non-European periphery; by the mid-1950s most new IBRD lending was going to Asia, Latin America, and (subsequently) post-colonial Africa. The pivot from reconstruction to development was not part of the institution's 1944 design; it was the operational adaptation to a context in which the original reconstruction problem was being addressed by another instrument.

The trade leg of the architecture proved the most fragile to assemble. The 1944 conference left trade institution-building to subsequent negotiation. The Havana Charter for an International Trade Organization, drafted at the UN Conference on Trade and Employment in Havana in 1947–1948 and signed in March 1948 by 53 countries, was an ambitious institutional design covering tariffs, quantitative restrictions, employment policy commitments, restrictive business practices, commodity agreements, and foreign investment. The Charter went to the US Senate for ratification and never came back; the Truman administration, facing Republican opposition that read the employment-policy and commodity-agreement provisions as commitments to economic planning the Senate would not accept, withdrew the Charter from consideration in December 1950. The ITO was abandoned. What survived was the General Agreement on Tariffs and Trade, signed in October 1947 by 23 founding contracting parties as a provisional instrument intended to lock in the tariff concessions negotiated at the first Geneva round pending ITO ratification. When the ITO failed, GATT remained: a narrower trade-rules regime, applied provisionally for forty-eight years until the WTO replaced it in 1995, that became the institutional core of multilateral trade liberalization by default. (§13.6 walks GATT in detail.)

The architectural-coherence claim is the diagram's interpretive content. Five operational elements were designed to interlock: current-account convertibility paired with permitted capital-account closure; fixed-but-adjustable parities under IMF surveillance; the gold-dollar standard with US commitment to dollar-gold convertibility at $35 per ounce; multilateral trade liberalization through GATT rounds; and long-term capital supplied by the IBRD. Trade liberalization through GATT depended on stable exchange rates so that tariff concessions retained their meaning across years of negotiation. Stable exchange rates required IMF balance-of-payments financing so that countries could sustain par values without being forced into the deflationary adjustment that the interwar gold standard had compelled. Deflationary adjustment could be avoided because capital-account closure protected domestic policy from speculative pressure that fixed parities and free capital mobility would have generated. The long-term reconstruction and development finance the IBRD supplied addressed the structural-investment demand that short-term IMF lending was not designed for. The system's premise was Keynesian in the institutional sense the previous section named: full-employment commitment as a domestic policy priority that international arrangements should support; capital-mobility skepticism as a design constraint rather than a temporary expedient; symmetric-adjustment intuition imperfectly realized in the asymmetric par-value settlement. The IMF's operational role as international central-banker analog, supplying international liquidity and coordinating exchange-rate management across national central banks, is the historical reference point for the modern central-banking debate that BQ06 develops.

The architecture's first major operational test came not from a balance-of-payments crisis but from a humanitarian and political emergency: the postwar reconstruction of Western Europe.

13.4 Le modèle de croissance endogène de Romer (1990)

The Marshall Plan was not primarily a fiscal-transfer program. It was institutional engineering on a continental scale, financed by US grants and loans whose direct fiscal contribution to recipient GDP averaged perhaps 2 to 2.5 percent at peak but whose conditioning, coordination requirements, and counterpart-fund mechanism worked changes on European economic governance the dollar amounts alone do not capture. The program ran from April 1948 through June 1952; total ECA disbursements across sixteen recipient countries were approximately $13.3 billion (roughly $150–170 billion in 2020 dollars). The institutional work was the larger story.

The instrument on the US side was the Economic Cooperation Administration (ECA), established under the Foreign Assistance Act of April 1948 and headed by Paul Hoffman. The ECA reported to the President but operated outside the State Department, with regional missions in each recipient country. The instrument on the European side was the Organisation for European Economic Co-operation (OEEC), established in April 1948 to coordinate the European response to the Marshall offer. Each recipient government submitted annual economic programs to the OEEC, which reviewed, allocated, and recommended disbursements; the ECA approved or modified them. The structure imposed multilateral coordination on countries whose economic policies had been substantially uncoordinated since 1939. The OEEC's progressive elimination of quantitative restrictions on intra-European trade ran in parallel with aid allocation, building what Alan Milward later named the European rescue of the nation-state, the practical pre-history of the European integration project the Treaty of Rome would formalize in 1957. (The OEEC reorganized as the OECD in 1961 with US and Canadian membership added.)

The conditioning had three formal dimensions plus a fourth through the productivity missions. First, intra-European trade liberalization through the OEEC's progressive elimination of quantitative restrictions, codified in the 1950 Code of Liberalisation. Second, national economic-program submission and review through the OEEC. Third, counterpart funds — the local-currency proceeds recipient governments collected when ECA-financed dollar imports were sold domestically — were held in special accounts and released only with joint US-recipient approval, providing continuing ECA influence on fiscal and investment decisions long after the dollar grants had been disbursed. In France, counterpart funds financed substantial portions of the Monnet Plan's industrial modernization. The productivity missions sent roughly 24,000 European managers, technicians, and trade-union leaders to US factories between 1948 and 1952, propagating a model of high-productivity, high-wage, mass-consumption industrial capitalism that the postwar settlement would substantially adopt.

The dollar-shortage problem of the late 1940s was addressed institutionally through the European Payments Union (EPU), established in July 1950 with Marshall Plan funding for the initial capital. The EPU was a multilateral clearing mechanism: bilateral surpluses and deficits among the seventeen European members were netted monthly and settled in EPU units (defined against the US dollar) rather than in scarce dollars themselves. Settlement was part-credit, part-gold: roughly 40 percent in gold or convertible currency and 60 percent in EPU credit in the early years, with the proportions shifting toward higher gold settlement as European currencies strengthened. The EPU was the workaround for the architecture's premature assumption that current-account convertibility would be quickly achievable; it allowed multilateral European trade to expand inside the dollar-shortage period without forcing each pair of countries into bilateral clearing. The EPU wound down in December 1958 when the major Western European currencies returned to current-account convertibility.

The distribution had a clear shape. The four largest recipients (United Kingdom, France, Italy, West Germany) absorbed roughly 65 percent of the total; adding the Netherlands brings the top five to nearly three-quarters. The concentration reflected both recipient-economy size and the political-strategic priority Washington attached to particular cases. Britain's substantial share, despite relative wartime industrial preservation, reflected the structural fact of British financial exhaustion. French aid funded Monnet's modernization investments. Italian aid was sized to the political contest with the Italian Communist Party that the 1948 election had crystallized. West German aid, channeled mostly as loans and starting later, built into the recovery that under Erhard's currency reform and the social-market framework would become the West German economic miracle of the 1950s. (The GDP map carries fuller long-run series for France, West Germany, and the United Kingdom.)

The Plan's structural-political effect was more substantial than its direct fiscal contribution. Bradford DeLong and Barry Eichengreen's (1993) reading, titled with characteristic compression “The Marshall Plan: history's most successful structural adjustment program,” treats the Plan as institutional engineering whose political-coordination payoff exceeded its modest GDP-share contribution. The OEEC framework built the multilateral-coordination habit that survived into the European Coal and Steel Community (1951) and the European Economic Community (1957). The conditioning entrenched the mixed-economy settlement that defined postwar Western European political economy: productivity-oriented industrial policy paired with welfare-state expansion, currency stabilization paired with capital controls, trade liberalization paired with sectoral protection. The productivity culture reframed labor-management relations around shared productivity gains rather than zero-sum distributional contest. The dollar transfers were the medium; the institutional coordination was the mechanism.

The Eastern Bloc was excluded by political process, not by design exclusion. Secretary of State George Marshall's Harvard speech of June 5, 1947 made the offer to all European countries; Soviet Foreign Minister Molotov attended the Paris foreign-ministers meeting later that month and departed after two days. Czechoslovakia and Poland initially indicated interest, then withdrew under direct Soviet pressure in early July 1947 — the Czechoslovak cabinet voted unanimously in favor on July 4 and reversed on July 10 after a delegation was summoned to Moscow. Stalin's calculation was that Marshall participation would have integrated the Eastern bloc into a Western-led order incompatible with the political consolidation Soviet policy was pursuing. The COMECON was established in January 1949 as the parallel institutional framework for the socialist bloc; ch. 15 picks up the parallel-systems story. The architecture's first decade was dominated by the dollar shortage and the Marshall Plan's political-coordination work; the second brought the first structural challenge.

13.5 The Order Operating, 1947–1971: Dollar Shortage to Triffin Dilemma

The Bretton Woods system as designed in 1944 did not begin operating as designed until 1958. The first fourteen years were spent solving the dollar-shortage problem the design had not anticipated. The next thirteen years were spent managing the structural contradiction that the architecture's success in restoring convertibility surfaced.

The dollar-shortage phase ran from the immediate postwar years through 1958. European and Japanese economies needed dollars to import US capital goods, food, raw materials, and oil; their exports to the United States were too small to earn dollars in the volumes required. The architecture's design assumption — that current-account convertibility would be operational shortly after the war with par values defended through IMF drawing rights and central-bank intervention — collided with the reality that there were not enough dollars in circulation outside the United States to make convertibility workable. Sterling's 1947 attempt at convertibility, mandated by the 1946 Anglo-American Loan Agreement, lasted six weeks before a run on sterling forced the British government to suspend it. The Marshall Plan's grants and the European Payments Union (§13.4) were the institutional responses; both bought time during which European production recovered and the dollar-shortage gap narrowed.

December 27, 1958 was the cusp. On that date eight Western European countries (Belgium, France, West Germany, Italy, Luxembourg, the Netherlands, Sweden, the United Kingdom) restored current-account convertibility for non-resident holders. Other European countries followed in the next several years; Japan's restoration came in 1964. The architecture as designed in 1944 began operating fourteen years later, and operated with substantial coherence for roughly the next decade. The IMF's par-value surveillance role became operationally meaningful in a way it had not been before; Fund lending expanded as members drew on quotas to defend par values against the larger currency flows convertibility enabled.

The dollar-shortage problem inverted into a dollar-overhang problem during the 1960s. The US current-account surplus narrowed as European and Japanese production recovered; the capital account moved into substantial deficit as American multinationals invested in European production capacity and overseas military spending accumulated dollar balances abroad. The cumulative US balance-of-payments deficit ran on the order of $1 to $3 billion a year through the early 1960s and accelerated thereafter. Foreign-held dollar balances, a fraction of US monetary gold reserves at the start of the 1950s, exceeded the US gold stock by 1960 and continued to grow as US gold reserves declined toward the gold cover the Federal Reserve maintained against domestic note issuance.

Robert Triffin, a Belgian-American economist at Yale, named the structural problem in testimony to the Joint Economic Committee of the US Congress in October 1959 and developed it in Gold and the Dollar Crisis, published in 1960. The Triffin dilemma identified an internal contradiction in the gold-dollar standard: the system required the United States to run sustained balance-of-payments deficits to supply the international liquidity that growing world trade demanded, but those deficits accumulated dollar liabilities against a fixed US gold stock at the $35-per-ounce parity, eroding the credibility of the dollar-gold convertibility commitment that anchored the par-value system. Either the United States stopped running deficits, in which case the world economy faced a liquidity squeeze; or the deficits continued, in which case dollar-gold convertibility eventually became unsustainable. The dilemma was structural, not policy-driven: no domestic-policy adjustment in the United States could resolve a contradiction the system's design had built in. (The formal model side of the open-economy Trilemma, which holds that capital mobility, fixed exchange rates, and monetary-policy autonomy form a trio of which a country can have at most two, lives in economics ch. 17; this chapter narrates the institutional history.)

The cooperative defenses against the dilemma's pressure built up through the 1960s. The Gold Pool, established in November 1961, was an arrangement among eight central banks (the United States, the United Kingdom, West Germany, France, Italy, Belgium, the Netherlands, Switzerland) to coordinate interventions in the London gold market and defend the $35-per-ounce price against private-market pressure. (The Gold Pool's coordinated central-bank intervention is part of BQ06's modern central-banking debate.) Bilateral Federal Reserve swap lines with major European central banks, expanded substantially after 1962, provided short-term liquidity in member-country currencies. The General Arrangements to Borrow (GAB), agreed in 1962 among the IMF and ten major members (the “Group of Ten”), supplemented Fund resources with standing credit lines. Special Drawing Rights (SDR), authorized at the IMF Annual Meetings in 1969 and first allocated in 1970, were a partial answer to the liquidity-supply problem the Triffin dilemma had identified: a new international reserve asset, allocated to member countries in proportion to their quotas. The first allocation totaled about $9.5 billion across three years, substantial but small relative to the dollar overhang the system was carrying by 1970.

Sterling's parity carried particular pressure. The pound was anchored at $2.80 from the 1949 devaluation through the 1960s; cumulative balance-of-payments deficits, sterling-area liabilities, and recurring confidence crises forced repeated defenses. The Wilson government, inheriting a substantial deficit in 1964, defended the parity through three years of successive runs before devaluing on November 18, 1967, to $2.40, a 14.3 percent reduction. The 1967 devaluation was the period's most consequential par-value change and signaled to markets that other major parities might be defensible only at substantial cost.

The terminal pressure built up in 1968–1971. US balance-of-payments deficits accelerated under the fiscal expansion that financed the Vietnam War and the Great Society without commensurate tax increases. The Gold Pool collapsed in March 1968 after sustained private-market gold buying overwhelmed its intervention capacity; the major central banks separated the official market (monetary gold trading among central banks at $35 per ounce) from a private market (gold trading at a premium that floated upward) into a two-tier system. The two-tier market preserved the par-value system's nominal anchor while abandoning its substantive defense against private pressure. The 1969 SDR creation came too late and at too small a scale to address the structural overhang. By 1970 the architecture was intact — convertibility operated, par values held with cooperative central-bank support — but the structural pressures were visible to anyone with access to the IMF balance-of-payments tables.

The system would close on August 15, 1971, when President Nixon suspended dollar-gold convertibility. The collapse and what followed are the next monetary chapter's subject. This chapter ends at the cusp, with the architecture intact and the structural pressures visible.

13.6 GATT and the Multilateral Trade Regime

GATT succeeded by being narrower than the institution it replaced. The Havana Charter for an International Trade Organization, signed in March 1948 by 53 countries, would have established a comprehensive trade-and-employment body covering tariffs, quantitative restrictions, restrictive business practices, commodity arrangements, foreign investment, and full-employment commitments. The US Senate refused to ratify; the Truman administration withdrew the Charter in December 1950; the ITO was abandoned. The General Agreement on Tariffs and Trade, signed October 30, 1947 in Geneva by 23 countries as a provisional step pending ITO ratification, became the trade-rules regime by default. GATT applied provisionally for forty-eight years until the WTO replaced it in 1995. Its narrower mandate — trade in manufactured goods with carve-outs for agriculture and developing-country industrialization — was the condition of its institutional survival.

Three structural commitments organized the regime. Most-favored-nation (MFN) treatment, codified in Article I, required that any tariff concession a contracting party extended to one trading partner be extended automatically and unconditionally to all others. MFN was the multilateralizing instrument: bilateral concessions became multilateral once each round closed, building progressive tariff reduction without requiring every pair of countries to negotiate every concession. National treatment (Article III) required that imported goods, once cleared through customs, receive treatment no less favorable than domestic goods on internal taxes and regulations — without it, countries could neutralize tariff concessions through internal regulation. Reciprocity structured each round: each country offered concessions roughly proportional to those it received, building the political case for ratification by ensuring domestic export interests gained foreign-market access in proportion to the import competition accepted.

The rounds ran on a roughly four-to-five-year cadence. Geneva 1947 was the founding round: 23 contracting parties, roughly 45,000 tariff concessions consolidated into a single multilateral schedule. Annecy 1949 added 11 parties (including West Germany through the occupation authorities). Torquay 1950–1951 expanded the membership further. Geneva 1956 ran a smaller round during European recovery. The Dillon Round (1960–1961) negotiated tariff adjustments occasioned by the formation of the European Economic Community; the EEC's common external tariff replaced the previously-separate national tariffs of its six member states, and Dillon-Round bargaining set the terms against non-EEC contracting parties. The Kennedy Round (1964–1967), authorized by the US Trade Expansion Act of 1962, was the period's largest. Sixty-two participating countries shifted from item-by-item bargaining to a linear-cut formula (a single percentage cut across broad tariff categories), achieved an average reduction of 35 to 50 percent on industrial-country manufactured goods, and concluded the regime's first non-tariff agreement (an Anti-Dumping Code).

The headline trajectory is the table's content. Average tariffs on industrial-country manufactured-goods imports fell from roughly 40 percent at the close of the Second World War (inherited from the 1930 tariff escalation and subsequent bilateral managed trade) to roughly 10 percent by the close of the Kennedy Round in 1967, the most successful sustained multilateral trade liberalization in history to that point. The mechanism was cumulative: each round locked in concessions through MFN extension, and the next round bargained from the previously-bound rates rather than from the higher pre-GATT levels. The Kennedy Round's linear-cut approach delivered a substantially larger single reduction than the item-by-item rounds had managed; linear cuts compress contested-item bargaining into a single across-the-board commitment harder to unwind through industry-specific objections.

The carve-outs were the regime's design, not its failure. Section 22 of the US Agricultural Adjustment Act authorized import quotas protecting domestic farm-support programs; a 1955 GATT waiver permanently exempted these from MFN discipline. The EEC's Common Agricultural Policy, established in 1962, built a structurally similar regime of variable levies and intervention buying that effectively excluded agricultural trade from GATT discipline. Article XVIII permitted developing countries to deploy quantitative restrictions, infant-industry tariffs, and balance-of-payments controls in support of import-substituting industrialization. The carve-outs were the political precondition for the manufactured-goods discipline that did get implemented: without them, the US Senate would not have ratified and most developing-country parties would not have signed. The regime functioned across two decades by being a manufactured-goods regime; the WTO from 1995 would attempt the agricultural extension GATT had structurally avoided.

The modern free-trade debate (the China shock literature, WTO accession effects, the protectionist turn of the 2010s) sits in BQ05; this chapter's GATT history is one input. Post-1971 trade-regime evolution (Tokyo Round, Uruguay Round and the WTO, the stalled Doha Round) belongs to a later chapter. What the architecture built across both monetary and trade dimensions is the synthesis section's subject; what it did not build is part of the same synthesis.

13.7 The Architectural Settlement: What Bretton Woods Built and What It Didn't

The Bretton Woods order produced a quarter-century of relative international monetary stability, recovery of European production, return to current-account convertibility, and sustained tariff reduction. It did so by building five things and not building four others.

What was built. A fixed-but-adjustable exchange-rate system on the gold-dollar standard, in which member countries declared par values, defended them through central-bank intervention with IMF financing, and could adjust them in cases of fundamental disequilibrium with Fund consent. An emergency-lending mechanism through the IMF that supplied conditional dollar liquidity to deficit countries, replacing the unconditioned and ultimately impossible adjustment that the interwar gold standard had demanded. A long-term-finance institution in the IBRD that supplied reconstruction capital to Western Europe in the late 1940s and pivoted to development finance through the 1950s and 1960s. A multilateral trade-rules regime in the GATT that reduced average industrial-country tariffs on manufactured goods from roughly 40 percent to roughly 10 percent across six rounds between 1947 and 1967. And the embedded-liberalism compromise — current-account convertibility paired with permitted capital-account closure — that protected domestic full-employment policy from the speculative pressure that fixed rates plus free capital mobility would have generated. The architectural diagram in §13.3 shows the institutional shape; the 1947–1971 operating record showed that the shape held.

What was not built. Keynes's true international currency was rejected; the bancor as an internationally created reserve asset settled across central-bank accounts was not the system the negotiators chose, and the dollar-as-anchor architecture that replaced it carried the structural contradiction Triffin would diagnose. Symmetric adjustment obligations on creditor countries were not built; the par-value system imposed conditions on deficit countries drawing on the Fund and faced no parallel obligation on persistent surplus countries to expand demand or revalue. A trade organization with the ITO's broader mandate — covering employment policy, restrictive business practices, commodity arrangements, foreign investment — was not built; the US Senate's refusal to ratify the 1948 Charter forced the GATT-as-rump settlement. And an institutional answer to the gold-dollar standard's contradiction was not built; the cooperative defenses of the 1960s bought time without resolving the problem at its source, and the SDR's 1969 creation came too late and at too small a scale to substitute for the dollar as primary reserve asset.

An interpretive frame in the secondary literature is worth naming. Eric Helleiner's States and the Reemergence of Global Finance (1994) reads the postwar order as the embedded-liberalism compromise John Ruggie (1982) had named: a deliberate settlement in which open international markets for trade coexisted with national capital controls so that domestic full-employment policy retained autonomy. Helleiner's contribution is the argument that the settlement was institutionally fragile: the capital-control regime depended on continuous political support, and the liberalization-of-finance pressures that built up through the 1960s (Eurodollar-market growth, multinational cash management, an emerging financial-services constituency) eroded the original coalition long before the formal collapse in 1971–1973. The chapter follows a more descriptive-institutional framing rather than adopting Helleiner's political-economy frame as the analytical spine; the frame is named as the most influential interpretive reading of the order's character.

The architectural-limit claim is the structural reading the chapter has carried throughout. The system's premise was finite by design: the United States supplied world liquidity by running balance-of-payments deficits, sustainable only as long as US gold stocks and the credibility of the $35-per-ounce convertibility commitment supported the dollar-as-reserve role. The Triffin dilemma was the formal expression of the limit. The cooperative defenses of the 1960s (the Gold Pool, the swap network, the GAB, the SDR) were management instruments for a structural problem that could not be management-resolved. By 1970 the architecture was intact and operating; the structural pressures were visible and accumulating. The Nixon shock of August 15, 1971, when the United States suspended dollar-gold convertibility, was the architecturally-determined endpoint of a system whose original design had not built an answer to the contradiction the design itself contained.

Two forward-links carry the chapter into its successors. The order's performance (the postwar golden age, decolonization, the import-substituting industrialization strategies pursued under Article XVIII flexibility) is the next chapter's subject; the architecture this chapter narrated is the institutional context within which that performance happened. The order's collapse (the Nixon shock of August 15, 1971, the Smithsonian Agreement, the move to floating rates in March 1973, the macroeconomic disorder of the 1970s) is the opening event of the chapter that follows; this chapter ends at the cusp, with the architectural pressures visible. (The moment Bretton Woods sits within the larger arc of doctrine entering global institutions is on the keynesian-era cluster of the intellectual-history timeline.)