Keynes and the International Origins of Macroeconomics

Keynes’s breadth as an economist and public intellectual has long haunted our understanding of the meaning and impact of Keynesianism in post-1945 political and economic history (Backhouse & Bateman, Reference Backhouse and Bateman2011: 21–45). For example, historians have debated the causal links between Keynes’s macroeconomic theory in The General Theory (Reference Keynes1936) and the ‘Keynesianism’ commonly associated with the economic policymaking of the USA and Western Europe after World War II. Historians of capitalism and political theorists, particularly, have considered Keynesianism as a phenomenon specific to the industrialised nation-state (Mann, Reference Mann2017). Echoing Marxian critiques (Sweezy, Reference Sweezy and Lekachman1963; Radice, Reference Radice1984), they have linked the concepts and policy recipes of Keynesianism to the political economy of imperial Britain. Indeed, a powerful critique suggests that Keynes’s supposedly universal recipe of fiscal expansionism greatly depended on access to cheap commodities and labour from imperial reservoirs (Biltoft, Reference Biltoft, Fredona and Reinert2018, Reference Biltoft2021; Goswami, Reference Goswami2018).

In this Element, I argue that limiting the significance of Keynes and Keynesianism to the industrialised nation-state is problematic. These readings overlook Keynes’s interactions with the League and its response to his ideas. Moreover, they overlook the UN’s significance. Thanks to its employment and development mandates, the UN and its multiple economic secretariats were an ideal platform for disseminating Keynesian concepts and toolkits across decolonising and postcolonial spaces.

The League and the UN were IGOs that embodied wide-ranging aspirations of realising world peace through international co-operation between nation-states (Sluga, Reference Sluga2013; Clavin & Sluga, Reference Clavin and Sluga2016; Kott, Reference Kott2024). The League contributed to the emerging development of a macroeconomic lens through its pioneering statistical compilations and support of business cycle research. This research documented a fragile international economic interdependence, composed of ‘national economies’ whose fluctuations could be measured and shaped through public intervention (Macekura, Reference Macekura, Unger, Ferns, Loveridge and Borowy2022: 137). It also rendered the League an outreach platform for interwar business cycle theorists whose contemporary importance often rivalled that of Keynes. Thus, we find Bertil Öhlin, from the planning-friendly Stockholm School, using the League as a space to define common policy ground with other more neoclassical thinkers such as Gottfried Haberler and Lionel Robbins (Endres & Fleming, Reference Endres and Fleming2002; Clavin, Reference Clavin2013). Keynes’s General Theory brought to the League’s openness to multiple theoretical viewpoints a self-contained yet conceptually powerful explanation of how, within the nation-state, output and employment were determined by the demand for capital and consumer goods.

When it comes to the post-World War II period, the existing literature on Keynes points in several directions. Although Keynes’s involvement in the Bretton Woods Conference (Reference Fleming and Moggridge1944), from which the IMF and the Bank emerged, is well known, in this Element I propose a longer temporal framework to understand the impact of his wartime projects on attempts to reinvent international economic governance. Substantial historical scholarship has also traced the roots of macroeconomics to the intersection between conflicting readings of Keynes (Reference Keynes1936) and the parallel mathematical formalisation of economics, which in the USA and Britain was accelerated by economists’s involvement in warfare production and its operational deployment.

Critically, the terminology in which Keynes dressed his macroeconomic analysis reoriented the accepted meanings of terms (such as ‘output’, ‘investment’, ‘saving’, and ‘aggregate/effective demand’) that quickly became indissociable from the language of post-1945 macroeconomics. The synthesis between Keynes’s macro lens and the utility maximisation of microeconomics was popularised, especially after Keynes’s death, by economists such as Paul Samuelson, Alvin Hansen, John Hicks, and Lawrence Klein. The historiography tells us that these thinkers offered a theoretical backdrop for policies like fiscal fine-tuning and deficit spending, which, although commonly described as ‘Keynesian’, were never endorsed by Keynes. Their macroeconomic outcomes were explained in classrooms through the IS–LM model, which ignored the psychological variables dear to Keynes and his disciples Joan Robinson and Nicholas Kaldor (Backhouse, Reference Backhouse, Backhouse and Bateman2006, Reference Backhouse2017, Reference Backhouse2024; Backhouse & Bateman, Reference Backhouse and Bateman2011: 63–112).

However, an overlooked problem tackled by this Element is how the post-war emergence of macroeconomics as a central branch of economics interacted with international economic thinking, as the League gave way to the UN in 1945–6. As we will see, not all macroeconomic thinking in the UN was Keynesian-derived, nor should the UN’s support of economic growth and development through macroeconomic control be labelled ‘Keynesian’. The UN pursued alternative macro lenses to Keynes’s, particularly Wassily Leontief’s input–output analysis (UN, 1960a; Tinn, Reference Tinn2023: 134). Instead, I argue that the secretariats of the UN’s economic bodies co-created, reproduced, and contested different types of Keynesianism – often contradictorily and through historically tenuous but performative assumptions about ‘what Keynes meant’. Following Backhouse (Reference Backhouse, Backhouse and Bateman2006: 20), to capture those varied understandings of Keynes’s thinking, I frame ‘Keynesianism’ and ‘Keynesian economics’ (I use both terms interchangeably) as a plural, interweaving arrangement of conceptual, policymaking, and rhetorical registers. These drew, even if often discursively only, from Keynes’s General Theory. By examining these registers, I aim to understand what difference Keynes’s analytical categories and Keynesian contributions to macroeconomics made to the UN’s applied work and to the legitimisation of its policy agendas.

Locating Keynesianism in the UN System

Whereas economic work in the League was concentrated in its Economic and Financial Organisation (EFO) and in the International Labour Organisation (ILO), the UN System comprises multiple economic agencies with overlapping tasks, which allowed for dynamics of collaboration and competition over the construction of macroeconomic knowledge and its policy applications. Those tasks included wide-ranging objectives such as raising living and employment standards, sustaining monetary stability, and supporting agricultural and industrial modernisation schemes towards ‘economic development’ (Unger, Reference Unger2018; McKenzie, Reference McKenzie2023; Selchow & Sluga, Reference Selchow, Sluga, Ferns and Villani2024). Attached to the UN General Assembly and its coordinating organ for economic and social matters, the Economic and Social Council (ECOSOC), are the Regional Economic Commissions for Latin America (ECLA), Europe (ECE), Asia (ECAFE), and Africa (ECA); as was (after 1964) the UN Conference on Trade and Development (UNCTAD), which attempted to redefine the international labour division between commodity producers and consumers. The UN System further encompasses specialised (and largely autonomous) agencies, such as the ILO, the Food and Agriculture Organisation (FAO), the IMF, and the Bank.

This Element examines how the secretariats of these economic bodies, often but not always in dialogue with academic economists acting as external advisors (Toye & Toye, 2004a: 211; Sluga, Reference Sluga2021), globalised Keynesianism under the three registers of concepts and toolkits, policymaking, and rhetoric. By concepts and toolkits, I mean the interaction between macroeconomic aggregates, national income accounting, and the forecasting of economic growth measured in Gross National Product (GNP) and Gross Domestic Product (GDP). Despite preceding efforts by Simon Kuznets and Colin Clark, the war meant that Keynes’s concepts heavily determined the national accounting methodology adopted by the UN with input from the Organisation for European Economic Co-operation (OEEC, later repurposed as the OECD). The state-building capacity of the UN System, which channelled economic expertise to postcolonial states via its ‘technical internationalism’, interweaved almost indistinguishably with the post-1945 consolidation of Keynesian concepts as the levers of macroeconomic analysis (Speich Chassé, Reference Chassé and Daniel2011, Reference Chassé, Daniel, Albert and Werron2020). To a degree beyond the League’s statistical work, macroeconomic aggregates gave to IGOs and their member-states substantial comparative and regulatory potential over the national economy (Sluga, Reference Sluga2013: 68). Models employing aggregates and hypotheses originating (or claiming to) from the General Theory supplied most of the macroeconomic knowledge, especially in the form of GDP projections, used by IGOs to tackle political objectives that went beyond the intellectual remit of Keynesianism: namely, commodity price stabilisation and international monetary reform.

The policy proposals generated by international secretariats did not always result in institutionally backed initiatives but were, rather, determined by power relations within the UN System and in relation to North and South member-states (Coats, Reference Coats1986). That is revealed by how Keynesianism intersected with international policymaking debates. The UN Charter’s (1945) foundational aim of ‘full employment’ was a prime objective of the General Theory. The idea of multilaterally enforcing ‘full employment’ was immediately pursued by UN officers and their external collaborators. Yet, during post-war negotiations of the stillborn International Trade Organisation (ITO), which was supposed to join the IMF and the Bank in multilateralising economic governance, commodity-producing countries framed ‘full employment’ as a policy specific to highly industrialised economies, unlike ‘economic development’. Analysing that particular history reveals how Keynesianism was also contested in and through international organisations.

A less obvious but powerful register through which international organisations engaged with Keynes and Keynesianism was by rhetorically evoking the historical Keynes to legitimise policy proposals. Because Keynes died in 1946, when the UN had barely begun functioning, we can generally only presume what Keynes might have thought about the work undertaken by those organisations, with an important exception being the issue of the IMF’s discretionary rights to interfere with national economic policy. Instead, we have abundant historical traces of how Keynes’s contemporaries (and subsequent generations of international officers) acted on contradictory understandings of his actions and ideas. That was partly because of Keynes’s shifting views on theory and policy, and partly because it took until the publication of Keynes’s Collected Writings in the 1970s for more rigorous historical narratives of his life to emerge (Bateman, Reference Bateman, Backhouse and Bateman2006). I will often refer to that rhetorical dimension as the ‘shadow of Keynes’, echoing an expression used by Elizabeth Johnson, the first editor of the Collected Writings, and her husband, the monetarist economist Harry Johnson. The Johnsons contributed to the 1970s backlash against Keynesianism and development economics by presenting their own historical critique of Keynes’s (supposedly) ‘distracting’ and overwhelming influence over post-1945 economics (Johnson & Johnson, Reference Johnson and Johnson1978: 228; Toye, Reference Toye1987: 22–31). Rather, by ‘shadow’, I mean an awareness of that thorny relationship between historical beliefs, ideas, and policy, which will inform my exploration of how rhetorical appeals to Keynes’s Clearing Union plan were central to UN thinkers’s justifications for multilateral projects of monetary and trade reform.

In that sense, the history of Keynes’s Clearing Union and its long run impact in the UN underscores how the histories of Keynesianism and IGOs are inseparable. They were traversed by a similar constituent tension between using economic expertise to strengthen national sovereignty and the awareness that raising the world’s living standards rested on multilateral economic governance. Although created by national will, the international-ness of IGOs gave their secretariats a breathing space to reimagine economic relations beyond the nation-state (Sluga, Reference Sluga2013). Keynes’s justification for multilateralising monetary and trade governance was a fallacy of composition like that sitting at the core of his macroeconomic theory, with its juxtaposition of general and particular interests: that the economic policies optimal for one (creditor) nation-state might be impossible for other (debtor) nations to pursue (Carabelli & Cedrini, Reference Carabelli and Cedrini2014: 121–8). That simple idea found an echo in the UN’s commitment to international co-operation and helped inform successive proposals emanating from UN thinkers to regulate the global economy.

The registers of concepts and toolkits, policymaking, and rhetoric are, foremost, a mapping guideline, rather than an analytical cage. Their boundaries are porous and their interarticulation reflective of agency, time and space. Following differing interpretations of the General Theory, Keynesian-derived or inspired toolkits were used to justify divergent policy options. For Hans Singer (Reference 82Singer and Singer1983), arguably the quintessential UN development economist, both Keynes’s and Keynesianism’s historical import amounted to the notion that expansionary aggregate demand was the engine of economic development. Jacques Polak, the legendary IMF Director of Economic Research, considered his own macroeconomic analysis of monetary demand to be Keynesian because it drew on the General Theory’s multiplier (Polak, Reference Polak2001). That said, Polak devised it to legitimise a lending conditionality that Keynes, as we shall observe, would have probably opposed.

How this Element is Structured

This study combines original archival research with scholarship from international, economic thought, and business histories to examine how the League of Nations and the UN System tackled Keynes’s thinking and contributed to the construction of post-1945 Keynesianism. My narrative begins in 1920 with the creation of the League. It ends in 1980 with the publication of the Brandt Report, which marked the end of the New International Economic Order (NIEO), and coincided with the beginning of the Latin American debt crisis. For conciseness alone, within the UN System, I will focus on the UN’s Department of Economic and Social Affairs (DEA/DESA), the UN’s Regional Economic Commissions, the ILO, UNCTAD, the IMF, and the World Bank. Because of their significant interactions with the UN, I make brief detours to the OEEC/OECD, private international organisations such as the International Chamber of Commerce (ICC), and financial bodies like the Bank for International Settlements (BIS). Finally, I use the terms Global North and Global South to refer to highly industrialised and commodity-producing countries, respectively, while being cognisant that those were non-homogeneous spaces, whose thinkers produced competing and overlapping notions of whether and how to potentiate national economic sovereignty through IGOs (O’Malley & Thakur, Reference O’Malley and Thakur2022).

Section 1 surveys Keynes’s interactions with the League to highlight the discontinuities caused by World War II in the history of international economic thinking. I explore how the EFO and the ILO used Keynes’s thinking as a foil for their work. The League, particularly, sponsored pioneering attempts to bridge Keynes’s macroeconomics with interwar business cycle theories, in ways that presaged the so-called ‘neoclassical synthesis’. I conclude by surveying the shadow of Keynes’s foundational moment: the 1944 Bretton Woods Conference.

Section 2 examines ‘full employment’ as a defining post-war Keynesian policy aim, and why, in the late 1940s, it was rebuffed by commodity-producing countries as perpetuating an unequal international division of labour that conflicted with ‘economic development’. I contextualise that discursive tension in the parallel creation of development economics and the constitution of the UN’s economic secretariat in New York. This history, I argue, helps explain why Keynesian concepts and toolkits proved fundamental to the UN’s development agenda, despite early critiques of their applicability to rural-dominated economies.

In Section 3, I investigate how macroeconomic aggregates became a standard analytical grid in the UN System through national income accounting and modelling toolkits. To achieve development by growth, UN bodies globalised Keynesian accounting identities and Harrod–Domar growth models. They also sponsored pioneering attempts at global macroeconometric modelling, such as Lawrence Klein’s Project LINK. Meanwhile, Polak and his IMF colleagues built open macroeconomic models with the opposing aim of contractionary macroeconomic control.

Section 4 explores how a discursive shadow of Keynes was pervasive in multilateral efforts to reform the Bretton Woods order following the ITO’s failure. Those efforts were not exclusive to Keynes’s disciples, although they were involved in UNCTAD’s foundational debates. More importantly, in the 1970s, UN thinkers used Keynes’s wartime writings to legitimise their commodity-stabilisation proposals, even when aware that Keynes’s own views differed from the UN’s political objectives. Finally, I explore why the Brandt Report was derided for its ‘Global Keynesianism’. The conclusion summarises my findings and discusses the complicated presence of Keynesianism in IGOs to this day.

1.1 The League of Nations and Monetary Reconstruction

After the Paris Peace Conference of Reference Keynes1919, Keynes hoped the League of Nations would be ‘an organ of justice and no longer of [national] interest’, and that it would neutralise the German reparations crisis through international cooperation (Reference Keynes1919: 138). When the USA decided against ratifying the Treaty of Versailles, and left Britain and France to quarrel with Germany over its war debt, Keynes reached a different conclusion: To escape the paralysis of national power politics, the League should limit itself to ‘the necessary organisation’ of issues uncontroversial to the ‘major powers’ (Keynes, Reference Keynes and Johnson1923b: 177; Markwell, Reference Markwell2006: 195–6). Independent of Keynes’s outlook, that objective did inform the League’s groundbreaking ‘technical internationalism’. The EFO and the ILO helped cement, at both national and international levels of economic policymaking, a technocratic belief in the efficiency of economic expertise, underpinned by statistical knowledge. Indeed, by using technical expertise to defuse politically fractious socio-economic issues, the League established a precedent for the UN System’s specialised economic agencies (Sayward, Reference Sayward2017: 7–20).

Keynes was acquainted with the work of the EFO and the ILO through his connections with former Treasury colleagues and academic contacts who joined the secretariats of these bodies. Among them was Walter Layton, the Liberal and future editor of The Economist, whom Keynes, due to the League’s dim political prospects, advised against becoming the first EFO Director (Hubback, Reference Hubback1985: 53–5). That said, the EFO was established to organise the first post-war financial summit: the 1920 Brussels Conference, which Keynes had helped convene (Clavin, Reference Clavin2013: 16–17). Keynes’s sceptical view of the League’s ability to shake off the weight of national interests was reciprocated by the EFO’s ambivalent responses to his own economic thinking. The EFO’s normative assumption of an international economy oiled by open capital markets and the fixed parities of a resurrected gold standard conflicted with Keynes’s reputation for monetary heterodoxy, especially following his critique of Britain’s return to gold in 1923. Reflecting a broad intellectual consensus among academic economists and central bankers, EFO economists saw in the gold standard as a means to combat post-war inflation, achieve political–economic stability, and restore international trade to pre-war levels (Endres & Fleming, Reference Endres and Fleming2002: 58–60; Mattei, Reference Mattei2022). That commitment informed its coordination of stabilisation loans to Austria, Bulgaria, and Greece, whose deflationary requisites soon made unpopular the EFO’s financial internationalism (Clavin, Reference Clavin2013: 16–34). The oversight functions inherent in the EFO’s monetary stabilisation plans have since been compared to the IMF’s surveillance of national capital accounts (Pauly, Reference Pauly1996; Martin, Reference Martin2022: 64–98).

Keynes’s view was the opposite. He feared that a generalised return to gold would unleash a strong deflationary shock into the world economy, as it eventually did in 1928 (Eichengreen, Reference Eichengreen1992). Keynes favoured pursuing domestic price stability through slightly fluctuating exchange rates, underpinned by Anglo–American central banking cooperation. His was an activist, yet predominantly bilateral, conception of monetary governance, at odds with the EFO’s gold internationalism (Keynes, Reference Keynes1923a: 35–6, 132–60). At the same time and in a way contrasting with his vocal critique of German reparations, Keynes (Reference Keynes and Johnson1923b: 176–7) kept silent over the political backlash caused by the EFO’s monetary intervention in Austria, deeming it an example of constructive international cooperation. Only in private did Keynes (Reference Keynes and Moggridge1928: 741–2) critique the League’s backing of the deflationary ‘rule requiring that a fixed percentage of gold be maintained against the note issue in spite of the fact that gold no longer circulates’.

The assumption that rekindling political–economic stability depended on restoring the gold standard permeated the 1920s EFO secretariat, including the sound money advocates who became key critics of Keynes’s macroeconomics, namely Per Jacobsson and Jacques Rueff. A rising figure in the French Treasury, Rueff worked at the EFO from 1926 to 1930 (Chivvis, Reference Chivvis2010). He became prominent for blaming British unemployment not on the gold standard (as Keynes) but, rather, on how unemployment allowances made wages sticky: a conclusion central to neoclassical economics that was opposed by Keynes and the ILO (Maurette, Reference Maurette1931). In 1929, on similar grounds, Rueff debated with Keynes the problem of German reparations at the Geneva-based Graduate Institute of International Studies, which was originally conceived as a training ground for League staff (Fleming, Reference Fleming2000). Jacobsson was a disciple of the Swedish economists Knut Wicksell and Gustav Cassel. He personified an alternative strand of Swedish monetary thinking to the fiscal interventionism popularised after the Depression by the Stockholm School. Jacobsson’s involvement in the 1922 Austrian reconstruction loan marked the beginning of a career bridging the BIS, the ICC, and the IMF – pulpits from which, as discussed further ahead, Jacobsson battled Keynes’s conceptions of international monetary governance (Jucker-Fleetwood, Reference Jucker-Fleetwood1979; Yago, Reference Yago2013).

In contrast to the EFO, the ILO’s advocacy of labour and social rights led its experts to engage more openly with Keynes’s critiques of the gold standard. ILO monetary and business cycle research insisted that bank credit management through monetary policy would safeguard internal price stability and employment. These objectives were made impossible by a return to gold (Bellerby, Reference Bellerby1924). At the centre of the ILO’s monetary understanding of business cycle fluctuations was the English economist John Rotherford Bellerby. He was trained at Leeds University and worked at the ILO from 1921 to 1927, from where he joined Cambridge and specialised in agricultural economics. Bellerby was ‘close to Keynes’ (Robertson, Reference Robertson1983: 408), who lauded his and the ILO’s advocacy of price-stabilisation via credit control (Keynes, Reference Keynes and Moggridge1924: 419).

Seen in the light of pre-war and interwar monetary thinking (Laidler, Reference Laidler1991, Reference Laidler1999), the ILO’s endorsement of credit control was hardly revolutionary. More remarkable is that, in parallel to Keynes, ILO researchers acknowledged the benefits of managed national monetary systems, based on fiat money and internationally supervised by central bank cooperation. Unlike their EFO counterparts, ILO experts rebuffed the idea that gold automatically regulated international exchanges and even admitted the benefits of sterilising gold inflows and adopting slightly flexible exchange rates. Keynes stressed those same ideas at an ILO conference on unemployment (ILO, 1924: 53–63). Keynes’s intellectual impact at the ILO, however, should not be exaggerated. Bellerby, for example, also benefitted in 1925 from a Harvard Fellowship, where he overlapped with Allyn Young. He was keen to learn about public works in the USA, anticipating ILO interest in economic planning throughout the Great Depression (Endres & Fleming, Reference Endres and Fleming2002: 18–27, 61–71; Who, 1981: 60). The 1930s would deepen the ILO’s engagement with Keynes – although, as I will show, he was only one ideational source amongst others.

1.2 International Responses to the Great Depression and Keynes’s ‘Innovations’

At the onset of the Great Depression, Keynes praised the EFO’s and the ILO’s ‘brave efforts’ to arrange for ‘exact quantitative information concerning contemporary economic transactions’ (Reference Keynes1930: 365–6). Despite their differences over monetary reconstruction, the EFO and the ILO shared an interest in assessing the ebbs and flows of an international economy increasingly riveted by protectionism. The EFO, especially, was enmeshed in a transnational network of research centres (the operation of which benefitted from Rockefeller Foundation support) aiming to statistically measure and predict the transition between cycles of expansion and recession, along with their transnational propagation (Cussó, Reference Cussó2012; Lenel, Reference Lenel2018; Ehrenfreund, Reference Ehrenfreund2023). Reflecting their institutional functions and the intellectual diversity of business cycle theories, the EFO’s and ILO’s staff disagreed over the economic benefits of fiscal-monetary interventionism. Accordingly, they reacted to Keynes’s macroeconomics (Reference Keynes1936) in diverse ways.

Although histories of the League and the UN System tend to stress the functional continuities bridging their economic and statistical work, examining how the EFO and the ILO engaged with Keynes uncovers significant differences. EFO and ILO collections of national and international statistics were pioneering in their transnational breadth but were not organised into common, overarching macro-categories resembling anything like Keynes’s aggregates. The EFO’s sponsorship of Jan Tinbergen’s research prefigured an enduring international commitment to econometrics, as examined in Section 3.2. Tinbergen’s endeavour, however, was met with scepticism not only by Keynes but within the EFO itself. In the following pages, I explore how the EFO’s and the ILO’s selective engagement with Keynes’s macroeconomics also hinted at the growing impossibility of ignoring the academic and discursive impact of macroeconomics over the language and categories used to conceptualise the national economy.

Once Britain left the gold standard in 1931 – followed by the USA in 1933 – and as the international economy fragmented into opposing trade and monetary blocs, the EFO looked to salvage monetary convertibility and expand trade. It banked on multilateral counter-cyclical responses to the Great Depression. This was particularly because from 1931 until its dissolution in 1946, the EFO’s Financial Section was directed by the British economist Alexander Loveday, who, despite subscribing to Rueff’s conclusion that unemployment allowances were aggravating the Depression, also sympathised with Keynes’s critique of the 1920s gold exchange standard (Clavin, Reference Clavin2013: 14–63; Ehrenfreund, Reference Ehrenfreund2023: 1178–9). Loveday saw the 1933 World Economic Conference in London as an opportunity to rewind trade protectionism through a new international monetary agreement that would reflate world prices. Under Loveday’s leadership, the EFO even suggested that Keynes should represent Britain in the Conference’s Economic Commission, a possibility rebuffed by the Prime Minister Neville Chamberlain’s scepticism about Keynes’s unorthodoxy (Clavin, Reference Clavin2013: 90).

Although unheeded at the World Economic Conference, Keynes’s multilateral scheme for world reflation in Means to Prosperity (Reference Keynes1933) received significant publicity. It explored themes he would revisit during World War II. Raúl Prebisch, then governor of Argentina’s central bank and still a sound money thinker (Caldentey & Vernengo, Reference Caldentey and Vernengo2016: 1727–9), acknowledged that Means’s critique of gold standard orthodoxy significantly impacted his intellectual trajectory (Calcagno, Reference Calcagno2021: 119). Keynes favoured a radical redistribution, supervised by the BIS, of US and French gold reserves to facilitate loan expenditure worldwide, thereby stimulating imports and relieving commodity producers. The breakdown in international cooperation, however, also motivated Keynes to explore alternative projects of imperial autarky through ‘national self-sufficiency’ (Skidelsky, Reference Skidelsky1992: 486–500; Nolan, Reference Nolan and Emmett2013), which were rejected by the EFO (Clavin, Reference Clavin2013: 132). Unlike the embryonic multilateralism of Means, Keynes’s take on national self-sufficiency continues to draw scholarly interest as exemplifying the economic dependencies hidden in his worldview (Biltoft, Reference Biltoft2021).

After the World Economic Conference, the EFO shifted attention to identifying the origins of and remedies for the Great Depression. The EFO’s Depression Delegation brought together leading (Western) economists to understand how business cycle downturns were transmitted, internationally, via monetary and price changes. Economists such as Alvin Hansen, Roy Harrod, and James Meade (an EFO official from 1938 to 1940) were at the forefront of integrating Keynes’s macroeconomics into neoclassical economics. At the EFO, they rubbed shoulders with critics and sceptics of Keynes: Dennis Robertson, Lionel Robbins, Charles Rist, Jacques Rueff, and Gottfried Haberler (De Marchi, Reference Marchi and Neil1991; Clavin, Reference Clavin2013: 202–5).

The Delegation discussed both Keynes’s General Theory and Haberler’s Prosperity and Depression. Published from 1936 to 1964 across several print runs, Prosperity was the most intellectually influential EFO report and a clear-cut example of how IGOs ‘technical work’ could shape national receptions of Keynesian economics (Sampaio, Reference Sampaio2016: 460). A proponent of an overcapitalisation explanation of slumps, Haberler surveyed an array of business cycle theories to advance a minimal counter-cyclical justification for monetary expansion and public works. His critique of the General Theory anticipated post-war neoclassical readings by portraying it as a special case in economics. Haberler believed that Keynes’s scenario of under-full employment equilibrium presumed the existence of sticky wages. By nuancing Keynes’s conclusion that wage cuts would reduce aggregate demand, Haberler echoed Hayekian and Rueffian claims that wage and price flexibility would always promote economic stability. That allowed him to relativise Keynes’s call for using fiscal policy to tackle unemployment (Boianovsky, Reference 70Boianovsky2000: 316–22; Boianovsky & Trautwein, Reference Boianovsky and Trautwein2006: 68–72).

Keynes corresponded with Haberler and questioned his attempts at minimising the General Theory’s theoretical novelty (Keynes, Reference Keynes1979: 266–75; Boianovsky & Trautwein, Reference Boianovsky and Trautwein2006: 49–52). On the one hand, Keynes claimed that he had radically reoriented economic analysis around a macro lens, determined by the interaction between aggregates: investment, saving, employment, and income. On the other hand, during the late 1930s, Keynes was not alone in providing a theoretical justification for counter-cyclical fiscal policy. Awareness of that broader intellectual conversation, of which the EFO’s Depression Delegation was a significant part, helps explain why Haberler and economists such as Bertil Öhlin dismissed Keynes’s claims to novelty as an example of Cambridge insularity. Haberler conceded that ‘thanks largely to Mr Keynes, there is to-day almost general agreement that Government spending, barring psychological repercussions […]will stimulate employment’ (Haberler, Reference Haberler1939: 236). Yet he also concluded that Keynes had mostly re-interpreted existing debates about the investment-saving relation through superfluous ‘terminological innovations’ (Reference Haberler1939: 218). Haberler’s barely disguised irony (Reference Haberler1939: 213–4) is evidence of the eventual break between the economic language and concepts structuring the League’s operational work and the UN’s.

Haberler’s Prosperity aimed to identify which economic variables were responsible for economic slumps. The EFO’s next step was to statistically test Haberler’s hypotheses (Clavin, Reference Clavin2013: 205–11). That task fell to Jan Tinbergen, who produced in 1938 the first macroeconometric model of the US economy (Maas, Reference Maas2014: 38–75; Dekker, Reference Dekker2021: 163–91, Reference Dekker2022). Amongst Tinbergen’s assistants was Jacques Polak, eventually the IMF’s Chief Economist from 1958 to 1979 (Boughton, Reference Boughton2011). Tinbergen’s groundbreaking work foregrounded the post-1945 ascendancy of econometrics, including within the UN System, as seen in Section 3.2. Nonetheless, the formal abstraction of Tinbergen’s research was met with scepticism from Loveday, who asked for Keynes’s feedback. In line with his psychological theory of investment, Keynes contended that unforeseeable events and individual expectations were hardly modellable, and that the coefficients Tinbergen took as given in his equations regularly changed in real life (Morgan, Reference Morgan1990: 101–25; Moggridge, Reference Moggridge1992: 621–4). Other heavyweight critics of Tinbergen’s work included Öhlin. But it was Keynes’s indifference that dismayed Tinbergen most. He considered that econometrics could help operationalise Keynes’s macroeconomic theory by determining the coefficients behind concepts and hypotheses fundamental to the General Theory – from the fiscal multiplier to the marginal propensity to consume (Ehrenfreund, Reference Ehrenfreund2023: 1185–6).

Keynes’s disillusionment with the practicality of Tinbergen’s research reflected his frustration with the League’s political inability to sanction Italy’s aggression against Ethiopia and Japan’s onslaught on China (Keynes, Reference Keynes1939: 568; Markwell, Reference Markwell2006: 197–201). Until Reference Keynes1939, the Depression Delegation still disagreed on the counter-cyclical benefits of cheap money and public works, and the EFO secretariat dismissed Keynes’s call (Reference Keynes1936: 378) for socialising investment as far-fetched (Clavin, Reference Clavin2013: 227). Whereas Keynes identified capital investment as the lever to rekindle demand in capitalist economies, another EFO initiative, under the guidance of the Australian Frank McDougall, surveyed the linkage between nutrition and living standards to stress the importance of boosting private consumption worldwide (Turnell, Reference Turnell2000).

Compared to the EFO’s struggles to define a common policy ground, the ILO secretariat welcomed with open arms the General Theory as a foil to the ILO’s endogenously produced counter-cyclical proposals. Historians have simplistically described this moment as the ILO’s ‘Keynesian’ turn (Maul, Reference Maul2019: 55, 85–91). However, even if Keynes was an important reference for ILO experts during the 1930s, he was not the only one. Harold Butler, a friend of Keynes and ILO Director-General from 1932 to 1938, took advantage of the Great Depression’s severity to expand the ILO’s economic research. Butler brought together experts from different intellectual traditions, including John Riches, a disciple of former EFO economist John Bell Condliffe; and the German econometrics pioneer Hans Staehle, who specialised in income inequality. Once the USA joined the ILO in 1934, the latter drew from the scientific networks surrounding Roosevelt’s New Deal administration, namely Lewis Lorwin, an expert in Soviet planning and advisor to Butler (Hughes & Haworth, Reference Hughes and Haworth2013; Plata-Stenger, Reference Plata-Stenger2020).

In the 1920s, the ILO staff had conceived of public works as procyclical (financeable only by budget surpluses, thus assuming balanced budgets). The Depression, however, led ILO experts to conclude, as Keynes eventually did, that monetary policy alone could not revive the world economy. Percival William Martin, who served in the ILO’s Scientific Division from 1920 to 1934, was a key figure behind that ideational turnaround. Cambridge-trained, Martin was unafraid of skirting the outer fringes of non-Marxian heterodoxy. While at Columbia University in 1929, he learned about Wesley Mitchell’s underconsumption theory and began explaining slumps as resulting from inadequate purchasing power. At the onset of the Depression, Martin placed greater emphasis than Keynes (who, at that time, still prioritised price stabilisation through credit expansion) on how fiscal policy could lessen output fluctuations (King, Reference King2017: 17).

Nevertheless, Martin and his colleagues extensively cited Keynes to back an international public works programme, discussed at the ILO 1932 Amsterdam Conference. At stake was Keynes’s argument that loan-financed public works were not inflationary and could revive hoarded savings (instead of crowding out private investment). To better conceptualise their public works proposals, the ILO secretariat corresponded in 1933 with Richard Kahn, Keynes’s closest disciple. Kahn was instrumental in refining Keynes’s (and the ILO’s) understanding of the fiscal multiplier, which, in turn, was essential to defining the expansionary limits of newly created public investment. Only by Reference Keynes and Moggridge1935 did the ILO start citing Stockholm School stalwarts such as Gunnar Myrdal, Bertil Öhlin, and Erik Lindahl, who were developing their Wicksellian-derived, rather than Keynesian, justifications for expansionary fiscal policy. Using those varied references, the ILO published, one year before Keynes’s General Theory, the report Public Works Policy (Reference Keynes and Moggridge1935), which offered an underconsumption rationale for loan-financed public works and used the multiplier to outline its limits (Endres & Fleming, Reference Endres and Fleming2002, 75–98).

The ILO secretariat was intellectually receptive to the General Theory’s ‘important bearing’ and published its first review, worldwide. The author was Abba Lerner (Reference Lerner1936: 435), the left-wing Keynesian economist behind ‘functional finance’, the framework of economic intervention opposed to sound finance. Keynes paid tribute to the ILO for standing ‘out conspicuously amongst the pronouncements of the numerous post-war international bodies’ and for its ‘consistent appreciation of this truth’ (Reference Keynes1936: 349). The ‘truth’ amounted to Keynes’s argument that the simultaneous pursuit of domestic expansionism, thanks to autonomous national monetary policy, would mollify international tensions by decreasing unemployment and expanding international trade. The problem was how to harmonise monetary and fiscal relations between competing national and imperial systems, an objective that would inform Keynes’s Reference Keynes and Harrod1943 International Clearing Union plan.

Before discussing the role of Bretton Woods in shaping the impact of Keynes in post-1945 international economic thinking, I will briefly examine another interwar experiment with financial internationalism: the BIS. Together with the ICC (Slobodian, Reference Slobodian2018), the BIS was a bridge between bankers, businessmen, and sound money economists, including Michael Heilperin, a foundational neoliberal who opposed Keynes’s economic interventionism. The BIS helped sustain an international commitment to capital liberalisation and anti-inflationary monetary policy, even after the slump, World War II, and the ideas of Keynes and like-minded thinkers had weakened the political acceptability of those tenets (Yago, Reference Yago2013: 38–108).

The key intellectual conduit between the League, the BIS, and the ICC was the former EFO economist Per Jacobsson. From 1931 until 1956, Jacobsson headed the BIS’s Monetary and Economic Department, after which he became IMF Managing Director until 1963 (Jucker-Fleetwood, Reference Jucker-Fleetwood1979). The BIS had been created in 1930 by Allied initiative as an ‘independent bank’ to depoliticise the collection of German reparations (Martin, Reference Martin2022: 99–133). After the 1932 Lausanne Conference froze reparations payments and rendered the BIS obsolete, it continued to function as a clearing house for gold, while also expanding Jacobsson’s research department. Keynes was sceptical of the BIS, not because of its technocratic functionality but because it served an orthodox conception of international monetary governance. He had hoped the BIS would join the League and the International Court of Justice as ‘a super-national authority of great influence which will provide an organisation for the collaboration of the central banks of the world’ (Reference Keynes and Johnson1929a: 335), for ‘the world […] to enjoy a rational monetary system’ (Reference 76Keynes and Johnson1929b: 345). This implied managing the price of gold to eliminate credit cycle fluctuations (Keynes, Reference Keynes1933; Markwell, Reference Markwell2006: 148–51), something at odds with Jacobsson’s views. Keynes recognised that under Jacobsson, the BIS became ‘the leading authority’ in banking statistics (Reference Keynes1934: 514), but in 1935, he dismissed BIS support for a return to fixed gold parities as the position of those ‘living in an unreal world, a fool’s world’ (Reference Keynes and Moggridge1935: 367).

The ‘fool’ was Jacobsson. Under his watch, BIS’s endorsement of a gold exchange standard continued after the 1933 World Economic Conference: Jacobsson knew Keynes personally and was unpersuaded by his advocacy for long-term low interest rates (Yago, Reference Yago2006: 5–8). When the BIS’s post-war survival was jeopardised by its acceptance of the Reichsbank’s looted gold, Jacobsson opposed Keynes’s Clearing Union and the Bretton Woods resolutions, which endorsed capital controls and called for liquidating the BIS (Toniolo & Clement, Reference Toniolo and Clement2005: 260–82). What ensured the BIS’s institutional survival was its status as an informal meeting place for European central bankers. The BIS’s claims of technocratic neutrality in service of international monetary stability set a precedent for the IMF’s governance and the critiques of political unaccountability the latter would generate (Daunton, Reference Daunton2023: 29).

After Keynes died in 1946, Jacobsson suggested that Keynes would have opposed those claiming to be Keynesian had he experienced post-war inflation (Jucker-Fleetwood, Reference Jucker-Fleetwood1979: 210). For Jacobsson, the ‘Keynesians’, amongst whom he included thinkers as diverse as Paul Samuelson and Nicholas Kaldor, advocated for cheap money and fiscal expansionism. Instead, Jacobsson argued that currency convertibility, rather than capital and price controls, was the optimal means for controlling inflation (Toniolo & Clement, Reference Toniolo and Clement2005: 288). By recalling Keynes’s distrust of inflation and balance-of-payments disequilibria as levers for boosting demand, Jacobsson concluded that the ‘Keynesians’ had betrayed their master (Yago, Reference Yago2013: 92–6). Here is the evidence that immediately after his death, Keynes’s name had already accumulated enough symbolic capital to be wielded by his opponents in international contexts. The background to Jacobsson’s evocation of the historical Keynes is unintelligible, however, without surveying the Bretton Woods origins of the ‘shadow of Keynes’.

1.3 Bretton Woods and the ‘Shadow of Keynes’

After the defeat of Poland in October 1939, Keynes admitted to Walter Layton that ‘I have never been so much of a League of Nations man as you. But my plan for the future would include the constitution of a European league on pretty drastic lines’ (quoted from Hubback, Reference Hubback1985: 168–9). Once Germany overran France in June 1940, Keynes joined the British Treasury in a wide-ranging advisory position, from which he helped construct an international, rather than a regional league of nations. Amongst his first tasks was countering Germany’s economic proposals for peacetime, namely Walter Funk’s Reference Keynes and Moggridge1941 New Order Plan. Keynes’s own plan, for an International Clearing Union, defined British positions ahead of the 1944 Bretton Woods Conference (Skidelsky, Reference Skidelsky2000: 76–81). It would include an International Clearing Bank that would manage a multilateral system of slightly adjustable exchange rates and capital controls, channel currency exchanges between central banks and, in turn, administer the first-ever supranational currency: Bancor. To prevent the gold sterilisation characteristic of French and US interwar monetary governance, Keynes supposed that Bancor would be convertible but not denominated in gold. Debtor countries would only be forced to hand in gold (or foreign exchange) to the Clearing Bank if they ran chronic balance-of-payments deficits (Reference Keynes and Moggridge1942a: 168–95; Moggridge, Reference Moggridge1992: 671–93).

The power imbalance between the USA and Britain meant, however, that the Bretton Woods Agreements generally reflected the US Treasury’s proposals of Harry Dexter White, rather than Keynes’s. Whereas the overdraft facilities of Keynes’s Clearing Union tallied $25 billion, the organisation that came into being, the IMF, was granted a kitty of $10 billion and its institutional arrangements were meant to safeguard the US’s creditor status (Skidelsky, Reference Skidelsky2000: 179–232; Boughton, Reference Boughton2022: 141–212). However, even though Keynes’s plan never saw daylight, it did endure in post-1945 international monetary and trade debates. Originally intended to act as the ‘pivot of the future economic government of the world’ (Keynes, Reference Keynes and Moggridge1942a: 189), the International Clearing Union’s popular topicality can be explained in part by its debtor-friendly structure and provisions for international investment and commodity price stabilisation. It was the culmination of Keynes’s twenty-year struggle to combine trade liberalisation and international monetary stability with the safeguarding of domestic employment.

Making his Clearing Union an automatic stabiliser of the international balance-of-payments reflected Keynes’s acute knowledge of Britain’s debtor position. Underlying his critique of discretionary powers and loan conditionality, was the belief that Britain’s loss of status made it susceptible of being treated worse than ‘the humblest and least responsible Balkan country’ (Reference Keynes and Moggridge1941: 46; Martin, Reference Martin2022: 219–43). Ahead of Bretton Woods, Keynes’s defence of economic interdependence was thus underpinned by a narrow assumption of which nation-states and empires should be in charge: the Commonwealth, the USA, major Western European countries, and the Soviet Union. As opposed to White’s concern with involving Latin American countries in the Conference (Helleiner, Reference Helleiner2014: 159–60), Keynes would have preferred an Anglo–American bilateral affair (Chandavarkar, Reference Chandavarkar1989: 113–33). He feared that, otherwise, it would become a ‘most monstrous monkey-house.’ On the back of this dubious image, he also complained that delegates from Brazil, China, and Mexico knew ‘little or nothing of international finance’ (Keynes, Reference Keynes and Moggridge1944b: 42–3).

How to explain, then, the enduring allure of Keynes’s Clearing Union plan to UN and Global South thinkers? Harry White’s pre-war support for an Inter–American bank made him realise, to a greater degree than Keynes, the value of a development agency like the World Bank (Boughton, Reference Boughton2022). Keynes’s plan, nonetheless, offered more capacious prospects for debtor countries than White’s. His Clearing Union conceived that up to 2 per cent of charges on surplus balances could be redirected, as Bancor, to finance an ‘International Investment or Development Corporation’ and ‘Commodity Controls’ (Proceedings, Reference Keynes and Moggridge1944b: 1569–71). In an unpublished 1942 memorandum, Keynes defined ‘Commodity Control’ as ‘a genuine organ of truly international government’, for ‘it is great facility not to have to ask for specific contributions from any named country, but to depend rather on the anonymous and impersonal aid of the system as a whole’ (Keynes, Reference Keynes and Moggridge1942a: 193). Although it is doubtful the Clearing Union would have possessed enough liquidity to undertake the tasks of the IMF, the Bank, and the stillborn ITO, it would be frequently evoked, from the 1950s, as a lost opportunity for creating an international reserve currency linked with development aid and commodity stabilisation (Cline, Reference Cline1976: 51–2).

Historical claims that Keynes remained oblivious to the economic problems of commodity-producing countries and their living conditions (Helleiner, Reference Helleiner2014: 222; Biltoft, Reference Biltoft2021: 28), or that his macroeconomic theory of full employment presupposed the exploitation of cheap colonial labour and resources, are thus too reductive. Instead, we must problematise the sharp contrast between the economic potential underlying Keynes’s proposals and the narrowness of their political underpinnings. Within its imperially inflected framework, Keynes’s Commodity Control acknowledged that income gaps between Southern producers and Northern consumers could be reduced by raising the former’s living standards through subsidies and by preventing wage dumping (Keynes, Reference Keynes and Moggridge1942b: 123–4). If Keynes’s proposal that commodity prices should reflect the conditions of the most efficient producers (as opposed to those with the lowest wages) implied a long-term decrease in world prices (Daunton, Reference Daunton2023: 182), he was not alone in reaching that conclusion. Both prioritising long-run efficiency and taking precautions against buffer stock schemes that would keep commodity prices artificially high were viewpoints shared by the League’s EFO (1945a: 266–8). Dropped after intra-governmental opposition, Keynes’s 1942 proposal for Commodity Control remained unpublished until rediscovered in 1973 by Lal Jayawardena, an UNCTAD collaborator (see 4.2).

Commodity stabilisation, in fact, took up only a small paragraph of the Clearing Union proposal officially published by the British government in 1943 (Moggridge, Reference Moggridge1992: 680–4). At Bretton Woods, Keynes was (successfully) pressed by the Mexican and future UN officer Víctor Urquidi to acknowledge equal support for reconstruction and development work by the eventual World Bank (Thornton, Reference Thornton2021: 89–97). Keynes’s vision for the Bank’s development agenda, however, did not go beyond raising the world’s living standards by making its resources ‘more fully available to all mankind’ (Proceedings, Reference Keynes and Moggridge1944a: 85; Grandi, Reference Grandi2017: 62). That amounted to intensifying the South’s economic dependence on the commodification of natural resources, a goal afterwards shared by Anglo–American ITO proposals, to the dismay of commodity-producing states intent on pursuing industrial modernisation programs.

Despite the complexities of Keynes’s post-war designs, contemporary critiques of his Clearing Union, on the one hand, and of the final Bretton Woods Agreements, on the other, illustrate how quickly Keynes became associated with the Bretton Woods order, while also being considered a source of inspiration for its reform. New York investment bankers and ICC-associated businessmen and economists vehemently opposed capital and exchange controls, attributing them to the pernicious influence of Keynes and the General Theory (Cortney, Reference Cortney1949; Helleiner, Reference Helleiner1994: 25–50). Sympathetic critics of Keynes included Raúl Prebisch, from 1948 the dynamo of the UN’s ECLA, and the Polish economist Michał Kalecki, later an intellectual star of the fledgling UN Department of Economic Affairs in New York. Although they concurred that Keynes’s Clearing Union was preferable to a US-dominated monetary order, Prebisch and Kalecki concluded that Keynes and White had erred in assuming a hypothetical international economic equilibrium. Peripheral countries, they argued, possessed asymmetric levels of monetary, economic, and financial autonomy, which generated structural balance-of-payments disequilibria. Those countries, Prebisch argued, needed expansionary full employment in the USA to sell their exports, combined with the ability to shield their infant industries through trade quotas, controls, and greater exchange-rate flexibility (Osiatyński & Toporowski, Reference Osiatyński and Toporowski2022: 82–5).

By prioritising monetary multilateralism over international trade reform, Bretton Woods left unsettled the conundrum of combining multilateral trade with the pursuit of import-substitution industrialisation in the South and reconstruction towards full employment in the North. China, India, and Mexico, along with smaller Southern American and Southeast Asian states, immediately criticised the meagre financial resources available to the IMF and the World Bank (Balasubramanian & Raghavan, Reference Balasubramanian and Raghavan2018; Daunton, Reference Daunton2023: 209–12, 307–16). Those unresolved tensions drove the 1946–8 negotiations of the ITO, the missing piece in the Bretton Woods order. As discussed in the following sections, the failure to create an ITO, which could have offered a powerful counterpoint to the IMF and the Bank, led UN thinkers, from Singer to Prebisch, to evoke Keynes’s Clearing Union as embodying a radically broader range of political–economic possibilities than the Bretton Woods Agreements. The assumption made by UN officers, namely Singer, that primary commodity prices faced a secular decline, if contrasting with Keynes’s own views, only reinforced attempts by international secretariats to legitimise their own contentious projects for monetary and trade reform by reference to Keynes’s Clearing Union. What lingered in their memory long after Keynes’s death were not the imperial boundaries of Keynes’s world vision, but rather his struggles to counterbalance US might and a dollarised world economy through pro-debtor financial multilateralism (Carabelli & Cedrini, Reference Carabelli and Cedrini2014: 123; Martin, Reference Martin2022: 220).

2.1 ‘Full Employment’ and ‘Development’ in the ITO

With memories of the Great Depression still fresh, full employment became a common post-World War II objective for Britain and the US. It was enshrined in the 1945 UN Charter and the 1948 UN Universal Declaration of Human Rights (Moyn, Reference Moyn2018: 82–107). There were non-Keynesian conceptualisations of what a national full employment political economy might be, namely, Kalecki’s (Mazurek, Reference Mazurek2023). Still, in international organisations, technical debates about how to define and reach full employment through public investment necessarily engaged with Keynes’s concepts, greatly because it was a quintessential policy objective of the General Theory and early Keynesian macroeconomics (ILO, 1946: 23, 326–42; Arndt, Reference Arndt1994). Keynes had sketched a theoretical path to full employment in industrialised economies with well-oiled monetary and fiscal apparatuses, besides the administrative expertise required to statistically conceptualise and manage aggregate demand. That said, there was also awareness in post-war trade negotiations that, with decolonisation looming, Keynesian full employment was not necessarily compatible with the national aspirations of undercapitalised, non-industrialised, and often still colonised commodity producers with underemployed labour reserves (Daunton, Reference Daunton2023: 235, 256–65).

The problem of aligning full employment in industrialised countries with industrial and agricultural expansion ‘elsewhere’ had been recognised in wartime League and ILO reports, as its experts had discussed how to expand post-war trade and improve living standards worldwide (Endres & Fleming, Reference Endres and Fleming2002: 198–236; Clavin, Reference Clavin2013: 285–337). Folke Hilgerdt (a trade and statistical specialist who joined the UN DEA as an Assistant Section Director) at the League, and Eugene Staley at the ILO, had concluded that import-substitution industrialisation in commodity-producing countries might create unemployment in industrialised economies. They remained, nevertheless, optimistic. North–South capital transfers to finance industrialisation and agricultural development schemes would expand world trade, in turn creating new export markets for North American and Western European businesses: provided that the North’s markets remained open and its wages flexible. Multilateral cooperation, Hilgerdt and Staley had concurred, could prevent a race to tariffs. Temporary commodity stabilisation schemes, including buffer stocks, could moderate global disequilibria in the production and trade of raw materials (League, 1945a, 1945b; Staley, Reference Staley1945). In the immediate post-war period, Christie Tait and Benjamin Higgins at the ILO agreed that although freer international trade remained the path to global prosperity, ‘capital-scarcity countries’ would be more concerned with ‘national development’ than ‘full employment’ (ILO, 1946: 25–6; Tait, Reference Tait1946).

The politically explosive task of rekindling world trade after two decades of mounting protectionism, while addressing global inequality, was among the ITO’s original aims. These negotiations began in 1946 and concluded at the Havana Conference of 1947–8. Had it been launched, the ITO would have joined and, at the very least, equalled the international standing of the IMF and the Bank (R. Toye, Reference Toye2003). Before negotiations began, the general assumption, shared by Keynes, was that full employment in the USA, along with reconstruction in Western Europe, would stimulate exports from primary commodity producers and their respective imports of capital and consumption goods from the North. As already mentioned, even Prebisch made the initial success of import-substitution strategies dependent upon full employment in the USA.

In the US State Department’s proposals for the ITO, full employment was deemed a core goal of the future organisation, but dependent upon substantially lowering trade barriers first. This was contested by the British and by commodity producers, if for different motives. Although Keynes had barely sketched his vision for the post-war trade order before his death, it partly foreshadowed the position of his disciple James Meade, a former League officer and one of the intellectual lights behind preliminary Anglo–American ITO talks (Schwob, Reference Schwob2007). Keynes called for controlled trade liberalisation by preserving Britain’s imperial preference and: ‘i) state trading for commodities; ii) international cartels for necessary manufactures; and iii) quantitative restrictions for non-essential manufactures’ (Reference Keynes and Harrod1943: 567–8). Meade’s position was more liberal and represented a third way between the US and that of wealthier commodity producers like Australia. Amidst post-war scarcity and inflation, Meade (Reference Meade and Howson1948: 83) acknowledged that full employment in Britain depended on stable commodity flows, imperial preference, and protecting British industries via quantitative restrictions.

That scenario implied preserving a structurally unequal division of labour between commodity producers and consumers. The latter, before and during the Havana Conference, from Australia and India to Ceylon, Brazil, and Mexico, worked intensively to shift the ITO Charter away from Anglo–American designs (Thornton, Reference Thornton2021: 121–44; McKenzie, Reference McKenzie2023: 95–128; Daunton, Reference Daunton2023: 262–5). Keynes (before dying) and Meade were sceptical, on grounds of political feasibility, of an ITO that would force the USA, its largest creditor member, to commit to full employment, increase imports and recycle its trade surpluses.

Yet the internationalisation of full employment was, since 1944, one of Australia’s aims for the ITO. Driven by the economist Herbert Coombs, Australian proposals took the political implications of Keynesian macroeconomics to their logical international extension. Coombs argued, on characteristic Keynesian grounds, that expanding world trade depended primarily on raising national income levels. Unlike US proposals, Coombs’s ‘positive approach’ to full employment made lowering trade barriers a consequence of previously expanding the world’s productive capacity at a national level. It assumed (like Keynes’s Clearing Union) that creditor surpluses should be recycled into imports and development loans to debtors (Turnell, Reference Turnell1999: 101–204; Rowse, Reference Rowse2002).

Despite its interest in raising commodity prices, Australia was a wealthy country compared to other commodity producers, one ruled by white settlers and politically sovereign, unlike China or India in 1946. Before the war drove commodity prices upwards, Latin American and Asian exporters had been deeply hit by the Depression. More than untrammelled trade liberalisation, they expected the ITO to let them access foreign markets while safeguarding the right to pursue national development programmes. More than an undefined ‘full employment’, they hoped the ITO would protect labour’s living standards from unregulated competition (Jensen, Reference Jensen, Jensen and Lichtenstein2016: 83–109; McKenzie, Reference McKenzie, Coppolaro and Mechi2020b).

Anticipating ideas often associated exclusively with Prebisch or ECLA, representatives from India, Ceylon, Colombia, and Mexico denounced the economic path-dependency inherent to the ITO’s twin-core aims of full employment and trade liberalisation. Rejecting a scenario of low-wage full employment underpinned by staple commodity exports, they insisted that the ITO should target the ‘development of economically backward areas and the raising of the standards of living of their peoples’ (UN, 1946a: 64). An Indian delegate participating in ITO negotiations was Palamadai S. Lokanathan, an LSE-trained economist and, from 1947 until 1956, the first Executive Secretary of ECAFE. Asserting that freer trade by itself would only conserve existing comparative advantages, Lokanathan stressed that India could not accept ‘that full employment in the major industrial countries alone would be enough to bring about full prosperity to other nations’ (UN, 1946a: 62). Arguing that raising effective demand with inadequate capital resources would aggravate the structural inequalities between commodity producers and consumers, Lokanathan concluded that the ‘whole idea of full employment’ was incompatible with India’s need to raise living standards by pursuing import-substitution and even send some of its population abroad. Not only would developing countries temporarily raise unemployment in industrialised nation-states, Lokanathan warned, but it was up to the latter to readjust their economies to ‘the requirements of the whole world’ (UN, 1946c: 21–2, 51–3).

Similarly, the Eastern European context of structural, hidden agricultural underemployment informed how Polish representatives critiqued Anglo–American ITO proposals. Such undercapitalised economies, cautioned the economist Tadeusz Łychowski, could not maintain ‘full employment’ and a ‘high level of effective demand’ (UN, 1946b: 2). Heading the Polish delegation to Havana was Stanisław Rączkowski, an economist who had written his doctoral thesis on the General Theory. Yet Łychowski’s references to hidden unemployment arguably reflected the foundational work of Kalecki, Ludwik Landau, and Rosenstein-Rodan, which anticipated post-war development economics (Mazurek, Reference Mazurek2023).

The introduction of protectionist carve-outs into the final Havana ITO Charter was meant to safeguard the conflicting goals of trade liberalisation, infant industrialisation, full employment, and imperial preference. It was a significant factor behind the Truman Administration’s refusal to submit the Charter to Congress for ratification (McKenzie, Reference McKenzie2023: 125–7). The ITO’s gap was filled by the General Agreement on Tariffs and Trade (GATT). Originally a legal agreement subsidiary to the ITO, the GATT never shook off its reputation as a vehicle for US-backed trade liberalisation, despite the efforts of its Executive Secretary Eric Wyndham White to make it an authoritative body (McKenzie, Reference McKenzie2020a). In these circumstances, before and beyond the 1964 UNCTAD, the legacies of Keynes’s Clearing Union and the ITO remained key junctures in an international narrative of discontent with the IMF-Bank duopoly over multilateral financial assistance and international lending.

2.2 The UN Secretariat at the Onset of Development Economics

While the ITO negotiations were underway in 1946, the head of the UN’s DEA, Arthur David Owen, began assembling his secretariat. The year after, the secretaries-general of the first UN Regional Economic Commissions, Gunnar Myrdal (for Europe) and Lokanathan (for Asia and the Far East) did likewise. While Anglo–American academia set the trend for post-1945 economics, and the General Theory helped change the language and the conceptual framing of economic problems at national and international levels, thinkers with their own macroeconomic theories, particularly Myrdal and Kalecki, joined the UN and co-defined its early economic agenda.

The following pages begin exploring how Keynesian economics shaped the UN’s work, despite critiques by early development theorists of Keynes’s ideas. After the war, Southeast and West Asian economists began questioning the General Theory’s practical relevance to what were contemporarily described as ‘backward’, ‘underdeveloped’, or ‘developing’ economies. One source of controversy was Keynes’s framing of the fiscal multiplier, which, by modelling the expansionary limits to public outlay, was essential to his definition of a full employment policy. Implicit in the multiplier was a monetised economy with well-defined markets for capital and consumer goods, whereas commodity-exporting countries were generally mono-economies with chronic balance-of-payments disequilibria, fragile consumer industries, and surplus agricultural labour (Okyar, Reference Okyar1951; Rao, Reference Rao1952a; Dasgupta, Reference Dasgupta1954). Indeed, if during the 1930s Keynes had worried about whether underpopulation menaced the survival of liberal capitalism, development economics, as operationalised within IGOs, had to contend with old Malthusian ghosts of overpopulation.

Yet despite the perceived limitations of Keynesian concepts and hypotheses, they became entrenched in the UN’s economic work and informed its embrace of macroeconomic regulation within a multilateral monetary and trade order. Drawing on the work of Albert Hirschman and John Toye, two development economists with an international background and, in their later lives, pioneering intellectual historians of development economics, I suggest that Keynes’s posthumous contribution to development theory and the UN’s economic work can be narrowed to two major aspects.

Firstly, the General Theory’s framing of capitalist economies as characterised by wage-price rigidities and/or operating at under-full employment equilibrium helped legitimise non-neoclassical theories of imperfect competition and structural disequilibria. Thus, Joan Robinson’s work on disguised unemployment, despite being known to historians as unrelated to Keynes’s work, helped entrench the perception amongst Cambridge alumni, such as Singer (Reference Singer1988: 37–40), that Keynesianism’s call for state intervention had something to say about rural underemployment. Speaking at the UN General Assembly in 1962, the Tunisian economist Chedly Ayari concluded that ‘after Keynes no one could deny the importance of population in economic theory or its importance to any development planner’ (UN, 1962a: 449).

Secondly, the widespread use of toolkits underpinned by Keynesian aggregates – from macroeconomic forecasting to growth models – rendered GDP a common measure to development economics and demand-management in industrialised countries (Hirschman, Reference Hirschman and Adelman1981: 49–73; Toye, Reference Toye2006a). To many self-described Keynesians, and even Latin American structuralists, Keynes offered an obvious starting point for economic planning under technocratic guidance, by justifying state action and macroeconomic management in the face of market failure. The Brazilian Celso Furtado acknowledged that before joining ECLA and working with Prebisch, Keynes’s ideas had given him a grasp of ‘the structural nature of the phenomenon of economic dependence’. By that, Furtado meant that ‘the capitalist economy cannot function without a certain centralisation of decisions [because] to some extent, capitalism is always State capitalism’ (Furtado, Reference Furtado1973: 34).

To illustrate the extent and limits of Keynes’s impact, I will explore the backgrounds of thinkers fundamental to the UN’s early work on employment and development. I look at individuals who identified themselves as ‘Keynesian’ or were simply trained in the new macroeconomics mainstream, and who either joined the UN secretariat (Kalecki, Sidney Dell, H.D. Fong, Hans Singer) or closely collaborated with it (V.K.R.V. Rao). It is worth noting that economic work at the UN was generally undertaken by the secretariat’s economists, statisticians, demographers, and even political theorists such as Walt Rostow, who was Gunnar Myrdal’s assistant from 1947 to 1951. The secretariat was supported by specially appointed expert committees, whose members were often sympathetic academics, picked to lend intellectual legitimacy to policy positions produced within the UN. In its foundational years, the UN’s centre of applied economic research was the New York-based DEA (reformed in 1955 as DESA). The DEA was expected to tackle economic questions through a global lens, by supporting the UN’s General Assembly and ECOSOC, including the latter’s subsidiary Sub-Commissions on Employment, Development, and Statistics. The DEA’s global lens was complemented by the Regional Economic Commissions for Europe (established in 1947); Asia and the Far East (1947); Latin America (1948); and Africa (1958) (Berthelot, Reference Berthelot2004; Fajardo, Reference Fajardo2022).

In its initial years, those UN bodies allowed economists to lead specialised agencies, create research departments from scratch, and work on topical problems at a moment when the political and technical horizons for radically raising global living standards seemed limitless (Kindleberger, Reference Kindleberger1955; Toye & Toye, 2004a: 66–7). Only at the IMF did economists hold greater influence over operational activities (Coats, Reference Coats1986; Chwieroth, Reference Chwieroth2010). Despite Leonard Rist’s efforts, project evaluation bringing together lawyers, engineers, and economists dominated the World Bank’s day-to-day work. It was in the mid 1960s that the Bank began assembling an institutionally strong economic division (Alacevich, Reference Alacevich2016; Hošman, Reference Hošman2023).

The inroads made by Keynesianism into early UN work were the product of recruitment networks that overlapped with those of rising Keynesian macroeconomists on both sides of the Atlantic, as well as younger economists’ openness towards state interventionism. H. D. Fong, the first Chinese official to join ECAFE, was hired by Lokanathan to lead its Research Department, where he stayed from 1949 to 1964. Fong had learned about Keynesian macroeconomics during his wartime exile in the USA, under the supervision of Alvin Hansen (Trescott, Reference Trescott2002). Key actors in the UN’s DEA and the Regional Economic Commissions also had direct or indirect links to Cambridge, where Keynes’s closer disciples were based (Richard Kahn, Joan Robinson, but also Richard Stone); to Oxford and the LSE (Roy Harrod, John Hicks, Rosenstein-Rodan, and James Meade, who traversed Cambridge and the LSE: Toye, Reference Toye2006a; Stewart & FitzGerald, Reference Stewart, FitzGerald and Cord2021; Saith, Reference Saith2022). Due to Britain’s imperial legacy, a significant number of Southeast Asian economists who joined the UN had been trained in Cambridge by Keynes’s disciples, and alumni ties to their former professors were reactivated at the UN (Orange-Leroy, Reference Orange-Leroy2024; Engerman, Reference Engerman2025).

Sidney Dell was an Oxford graduate and arguably the foremost UN economic expert until the 1980s. His long career began at the DEA (where he seconded David Owen) and spanned ECLA, UNCTAD, and the UN Development Programme. Upon joining the UN in 1947, Dell learned from Kalecki, Singer, and Kaldor, who remained one of his closest friends (Abrahamian, Reference Abrahamian, Helleiner, Abrahamian, Bacha, Lawrence and Malan1995: 3–30). Singer, a Cambridge alumnus, was recruited to the DEA after overlapping with David Owen at Glasgow University and in government work. Escaping what he considered to be the LSE’s stifling atmosphere, Kaldor became, from 1947 to 1949, the first Research Director of the ECE’s Planning Department. Kaldor had been recommended to Myrdal by Rosenstein-Rodan (Thirlwall, Reference Thirlwall1987: 104–11; Shaw, Reference Shaw2002: 34–40; Toye & Toye, 2004a: 56).

Academic networks, however, did not always facilitate the transmission of Keynesian macroeconomics into UN settings, as seen from the case of Keynes’s former student, Vijayendra K. R. V. Rao. Together with Singer and Alec Cairncross (an OEEC economist), Rao was among Cambridge’s first PhDs in Economics. His thesis on Indian national income was supervised by Colin Clark (Rao, Reference Rao2002: 50–5). Rao later directed the Delhi School of Economics and, as a national expert collaborating with ECOSOC, lobbied with Singer in the early 1950s to equip the UN with a soft-loan fund (the SUNFED). Although he admired Keynes, Rao questioned the relevance of demand analysis and the multiplier to India. Instead, and like the Saint Lucian, LSE-trained economist Arthur Lewis, Rao recovered classical political economy’s notion of raising labour productivity and capital accumulation: thrift and hard work were the commonsense priorities for developing countries (Rao, Reference Rao1952b: 67). Rao’s and Lewis’s use of assumptions contextual to Modern England are not only reminders of the multiple historical geographies underlying early development economics but also help us understand why the association of Keynes’s thinking with Anglo–American backgrounds did not preclude Keynesian macroeconomics from being used in very different contexts.

Comparing the international careers of Singer, Kalecki, and Kaldor further demonstrates the layered influence of Keynesianism in how early UN economists and experts tackled the problems of development and employment. Singer is usually described as the definitive Keynesian development economist within the UN. Still, Singer’s indebtedness to Keynes and Keynesianism related chiefly to his choice of analytical tools (the macroeconomic framework applied to demand analysis through national income accounting and growth models) and broadly comparable political visions about international monetary and trade reform (Toye, Reference Toye2006b). By contrast, it was Singer’s concern for the distributive justice of international trade, a matter in which Keynes had shown little interest, that guided him towards the argument that the barter terms of trade of primary commodity exports faced a secular decline against industrial manufactures (UN, 1949a). Prebisch, who arrived at similar conclusions independently, then politically repackaged Singer’s findings (UN, 1949b; Toye et al., Reference Toye2003).

The resulting Singer–Prebisch thesis underpinned ECLA structuralism and gave formal backing to commodity exporters’s calls for import-substitution industrialisation and trade reform. Singer broke away from over a century of classical and neoclassical understandings of international trade, which assumed that in a Malthusian overpopulation scenario, the prices of primary products would continue to rise, an assumption shared by the young Keynes (Toye, Reference Toye2000: 219–26). Nevertheless, and despite what separated his work on the applied economics of development, welfare, and health from Keynes’s focus on industrialised economies and high-level policymaking, Singer consistently named Keynes (alongside Joseph Schumpeter) as his main intellectual influence. Following his own historical understanding of Keynes, Singer repeatedly interpreted the history of international economic reform, from his terms of trade thesis to the creation of UNCTAD, as directly rooted in Keynes’s vision (Singer, Reference Singer1984).

While overt assumptions about the meaning of Keynesianism permeated Singer’s narratives about the origins and purpose of his work, Kalecki’s international trajectory shows how a non-Keynesian could implicitly use Keynesian aggregates to convey his ideas about inflation and development. Kalecki joined the ILO from Oxford in March 1945. During the 1930s, slightly before Keynes, he had built on Marxian underpinnings his own macroeconomic theory explaining the short-term determinants of output and employment. Kalecki’s ILO reports assessed the difficulties of upholding full employment without inflation in economies affected by supply-side disruptions, a theme previously pursued by Keynes’s Reference Keynes1940 How to Pay the War. Yet the intellectual animus guiding Kalecki’s post-war analyses of full employment was not Keynesian. His search for a non-inflationary full employment policy reflected as much the experience of food shortages in wartime Britain (Toye & Toye, 2004a: 72), as the agricultural unemployment of interwar Poland (Mazurek, Reference Mazurek2023). Further, Kalecki’s attention to class conflict led him to critique Keynes by arguing that full employment would be opposed by businessmen striving to use unemployment to discipline labour (Toporowski, Reference Toporowski2018: 110–25).

Restless at the ILO, Kalecki joined the UN’s DEA in 1948 to head its Economic Stability Division. There he began deploying his macroeconomic theory to analyse the problems of employment and development, eventually building upon a macroeconomic gridline that ultimately derived from Keynes’s General Theory (Toye, Reference Toye2006a: 987). That also meant serving as a friendly tutor to younger colleagues such as Dell and Singer. Ahead of his peers, Kalecki integrated the problems of increasing returns and imperfect markets, with an account of how misguided foreign investment could lead to an unbalanced expansion of the industrial sector, generating food shortages and potentially undermining development projects (Toporowski, Reference Toporowski2018: 173–82; Osiatyński & Toporowski, Reference Osiatyński and Toporowski2022: 17–43).

Kalecki, however, only indirectly addressed the problem of development at the UN. His chief task was identifying the international conditions and obstacles to full employment during the post-war inflationary context and through the Korean War. From 1947 to 1954, Kalecki co-edited the UN’s World Economic Report, the spiritual successor of the League’s World Economic Survey. With a talented and diverse staff (comprising Keynesians like Éprime Eshag and critics of Keynesianism like Dudley Seers), Kalecki’s team based their country analyses on a deconstructed Keynesian analytical grid, which started from GNP and analysed the causes behind yearly changes in key aggregates such as investment and consumption. On that statistical canvas, Kalecki deployed his sophisticated awareness of the structural imbalances between agricultural and industrial sectors, or of the problems in regulating consumption between capitalists and workers in open but undercapitalised economies (Dell, Reference Dell1977; Toye & Toye, 2004a: 71–5).

Kalecki’s duties also included sitting on ECOSOC’s Sub-Commission on Employment and Economic Stability. From his dialogue with invited experts, including the econometrics pioneer Ragnar Frisch and Keynes’s biographer Roy Harrod, the Sub-Commission concluded in 1949 that without multilateral coordination, UN member-states were likely to restrict imports to safeguard employment levels. Because the IMF lacked the resources to prevent a major recession, the Sub-Commission feared the consequences for world trade expansion. That fear, as I will explore, motivated the Sub-Commission to launch the UN’s only attempt at enforcing an international full employment policy (Toye & Toye, 2004a: 87–91; Toporowski, Reference Toporowski2018: 171–3).

2.3 ‘Extreme Keynesianism’: The End of Full Employment as a UN Goal

In 1949, the Sub-Commission on Employment greenlighted an expert group tasked with defining ‘National and International Measures for Full Employment’ (hereby referred to as NIFE). The resulting report represented the apogee of full employment as a UN policy target. NIFE was also Kaldor’s most notable contribution to UN thinking, despite his subsequent contributions to endogenous growth models and development theory. After leaving the ECE for Cambridge in 1949, Kaldor agreed to head the NIFE group of experts, which included Pierre Uri, who had helped introduce Keynesian national income accounting in France (Sampaio, Reference Sampaio2016: 494–9).

Although he received assistance from Kalecki’s team, NIFE was fundamentally the product of Kaldor’s pen. The report framed full employment as a multilateral goal essential to international economic stability, which because it depended on some ‘kind of government action’, Kaldor deemed common to capitalist and socialist systems alike (UN, 1949c: 7). NIFE repackaged Keynes’s (and Coombs’s) theory that domestic fiscal expansion in industrialised countries would help commodity producers by expanding trade, international lending, and direct foreign investment (Turnell & Ussher, Reference Turnell and Ussher2009). Although acknowledging that commodity-producing countries relied on an expanding world economy, NIFE was unconcerned with the agricultural underemployment and capital scarcity characterising those economies. Kaldor believed those problems belonged to the realm of economic development and framed ‘full employment’ as a policy meant for industrialised countries, which centred on tackling (industrial) unemployment caused by insufficient effective demand. The international pursuit of full employment, in Kaldor’s view, thus ran parallel to the separate task of assisting ‘the underdeveloped countries’ in fighting disguised unemployment through capital investment, under the assumption that labour could be shifted to industry without decreasing agricultural output (UN, 1949c: 12; Rao, Reference Rao1952b: 49–50).

Like Keynes’s Clearing Union, NIFE shouldered the USA, the world’s leading creditor, with the ‘burden’ of global economic adjustment. Kaldor’s practical proposals included stabilising commodity prices; using the IMF to recycle creditor surpluses into foreign investment; and raising the World Bank’s lending capacity. ECOSOC would coordinate national commitment to internationally agreed full employment targets and define adjustment priorities for surplus and debtor countries (UN, 1949c: 49–69). In response, the IMF and the Bank stonewalled NIFE to protect their jurisdictions from UN interference (Toye & Toye, 2004a: 98–9). Highly influential economists in US academia, namely Charles Kindleberger and Jacob Viner, attacked NIFE as pro-inflationary. Kindleberger (Reference Kindleberger1955: 346) demolished NIFE as the ‘high-water mark in academic naiveté’ of the UN secretariat: a ‘fairly extreme Keynesian solution was […] presented to a world which was not prepared to accept extreme Keynesian solutions (even though Keynesian analysis has been widely adopted)’.

Such dismissiveness came in the wake of McCarthyism, which had a devastating impact on the morale of the UN secretariat and which, in 1951, had prompted the dissolution of the Sub-Commission on Employment and Economic Stability. NIFE’s redistributive economic policies would have committed the US to undertake significant macroeconomic adjustments, rather than more limited development aid disbursements (Toye & Toye, 2004a: 99–109). To protect the UN from US criticism, Trygve Lie’s successor as Secretary-General, Dag Hammarskjöld, began from 1953 to position the UN’s international economic work as technical and apolitical, thereby rebuffing NIFE and the wider ‘Keynesian’ bent of the UN’s DEA (Toye, Reference Toye2006c). In 1955, Sidney Dell told Kaldor that ‘we are told from on high […] that to discuss economic trends in aggregative terms is “too Keynesian”’ (quoted by Toye & Toye, 2004a: 101).

Yet, before Hammarskjöld joined the UN, a follow-up report to NIFE had been published, which significantly helped shift the DEA’s attention away from employment problems and towards development, while facilitating the diffusion of Keynesian ideas nonetheless. The report was mostly drafted by Arthur Lewis and targeted agricultural underemployment (UN, 1951). It advanced an early version of Lewis’s Nobel Memorial Prize-winning growth model, which, unlike the Harrod–Domar model analysed in Section 3.2, justified the benefits of capital accumulation on classical, rather than Keynesian grounds. Lewis argued that surplus agricultural labour could be shifted into newly created export industries. Given a mismatch between overpopulation and capital scarcity, the marginal productivity of labour in agrarian economies lagged far behind that of capital (Tignor, Reference Tignor2006: 79–98). Lewis’s model was a significant and non-Keynesian contribution to development economics’s concern with using agricultural and industrial modernisation, via capital investment, against underemployment (Arndt, Reference Arndt1987: 49–60).

In the 1954 article that won him the Nobel Prize, Lewis dismissed the General Theory as a special case of neoclassical economics. According to Lewis (Reference Lewis1954: 139–40), Keynes assumed the existence of unlimited supplies of labour, land, and capital that remained inaccessible only because of insufficient effective demand, as in industrialised economies. Yet Lewis’s defence of state planning in his UN report indirectly facilitated the international dissemination of Keynesian economics. By backing technocratic planning at national scales, Lewis reinforced ongoing work by the UN’s multiple economic agencies to measure and evaluate economic performance across North and South through growth rates measured in GDP (Toye & Toye, Reference Turnell2004b).

As explained in Section 3, capital-output and investment ratios became the policy targets of both reconstruction and development, articulated through Keynesian toolkits like Harrod–Domar, which relied on a national income accounting framework that originally reflected the specificities of the industrialised nation-state. From the 1950s to the 1970s, international work specifically on employment in industrialised and commodity-producing countries continued at the ILO. To adapt its full employment mandate to ‘developing’ countries, the ILO spent the first post-war decades rethinking its conceptualisation of unemployment to account for informal labour (Benanav, Reference Benanav2019). That story too, however is interwoven with the Keynesian metrics of economic growth.

3.1 The Contested Globalisation of the UN System of National Accounts

Until World War II, economic statistical benchmarks were primarily financial (balance-of-payments), budgetary (fiscal surpluses and deficits), monetary (gold reserves and price levels), and international trade-related (imports and exports of goods) (Schmelzer, Reference Schmelzer2016: 75–102). Statistical yearbooks such as the League’s World Economic Survey, however pioneering in their global scope and epistemic logistics (Bemmann, Reference Bemmann, Almada, Santos and Santos2025), lacked an overarching statistical framework such as GNP/GDP. This was true – even for the pre-war editions edited, from 1937 to 1939, by Keynes’s disciple James Meade, who introduced a chapter surveying public finance and national income trends (Mitra-Kahn, Reference Mitra-Kahn2011: 219; Howson, Reference Howson2026: 146–8). As the UN absorbed the League’s economic functions, amongst its priorities was statistical harmonisation at national and intergovernmental levels, including but not limited to national income accounts. By facilitating cross-country comparisons of GNP’s components, from consumer expenditure to capital formation (UN, 1948), an internationally standardised accounting system would facilitate IGOs’s supervision of and assistance to member-states (Ward, Reference Ward2004: 36–76).

Both the dying League and the rising UN heavily relied on Anglo–American expertise. Among the League’s last pre-war initiatives was a vague resolution from its Committee of Statistical Experts (1939) to harmonise national income accounts; the Committee gave comparatively more attention to improving balance-of-payments statistics. Meanwhile, Colin Clark in Britain, and Wesley Mitchell and Simon Kuznets in the USA, had developed pioneering national income accounting techniques that excluded final government expenditure from their computations. They assumed that military spending was ultimately financed by the fruits of private initiative and negatively impacted public welfare (Suzuki, Reference Suzuki2003).

Keynes’s How to Pay for War (Reference Keynes1940) departed from those preceding contributions by integrating public and private expenditure into a definition of Gross National Income (GNI) that included war-related output. How to Pay was also the ‘first book’ in economics read by a ‘14-year-old’ Angus Maddison (Reference Maddison2013: 124), who, in 1953, joined the OEEC. There, Maddison would become a seminal scholar of quantitative macroeconomic history, renowned for his national income databases that attempted to chart multisecular economic growth.

In applying his macroeconomics to the problem of controlling inflation in a full employment war economy, Keynes provided the conceptual foundation for the work of another key economist, Richard Stone. Austin Robinson hired Stone and James Meade at the War Cabinet Office to prepare a comprehensive set of British accounts using Keynes’s definition of GNI. In 1941, under Keynes’s supervision, this work was published as a White Paper. In turn, US economists Richard and Milton Gilbert, followed by Nancy and Richard Ruggles, adopted a similar definition of GNP that benefitted from Keynes’s feedback (Mitra-Kahn, Reference Mitra-Kahn2011: 210–73). Stone and the Ruggles would soon collaborate through the OECD and the UN.

In 1945, Stone was invited by Alexander Loveday to chair a Subcommittee on National Income Statistics of the League’s Commission of Statistical Experts, whose work had been interrupted in 1939. The conceptual detail of Stone’s proposed social system of accounts prevented its widespread adoption (Schmelzer, Reference Schmelzer2016: 93–4). Still, the UN Statistical Office immediately adopted its basic ‘definitions and methods’ (1947b: 2). Although avowedly reproducing ‘the model of an advanced industrial economy in which transactions in money are dominant’ (UN, 1947a: 24), Stone’s framework underpinned the UN’s first compilations of National Income Statistics of Various Countries (UN, 1950: 4).

Stone also became, in 1947, the first chairman of the International Association for Research on Income and Wealth, whose founding members included Jan Tinbergen and Nancy and Richard Ruggles (Vanoli, Reference Vanoli2005: 427). After being denied a position at Yale, Nancy Ruggles joined the UN and was Assistant Director of its Statistical Office from 1975 to 1980 (Ruggles & Ruggles, Reference Ruggles and Ruggles1999). If the Association remained an influential site for academic exchange and the UN became the eventual globaliser of national income accounting, it was the West European-focused OEEC that effectively kick-started its international standardisation. Created to supervise the disbursement of Marshall Plan funds, the OEEC’s reconstruction task was hampered by post-war inflation. Keynes (Reference Keynes1940) had shown how national income accounting could be marshalled to forecast the fiscal retrenchment or stimulus needed to combine employment with price stability. Working at the US State Department’s European Cooperation Administration, the Ruggles convinced Stone to collaborate with the OEEC, whose first Secretary-General was the French economist Robert Marjolin (Schmelzer, Reference Schmelzer2016: 94–110).

The result of all this international activity was that macroeconomic aggregates became the statistical concepts by which IGOs measured ‘national income’, thus making it easier for the meaning of expressions like ‘investment’, ‘consumption’, and ‘saving’ to be re-interpreted around Keynes’s propositions in the General Theory (Backhouse, Reference Backhouse2024: 373). During the 1950s, the OEEC accounting system (Reference Keynes and Harrod1951) was repurposed, again under Stone’s direction, into the UN System of National Accounts (1953). It articulated the same Keynesian ‘interrelationship between production, consumption, and investment’ (Ward, Reference Ward2004: 77), whilst responding to the UN’s search for a ‘simplified system of accounts with a smaller number of entries’ (UN, 1950: 5). By restricting economic life to measurable monetary flows, the SNA overlooked barter trade, unpaid and informal labour. Poorer European states and colonial and postcolonial states often lacked the political centralisation and bureaucratic capacity to quantify categories which, in the first place, were ill-fitted to their societies (Schmelzer, Reference Schmelzer2016; Macekura, Reference Macekura2020). It was only in the late 1950s that the UN and the OEEC began publishing comprehensive and real (rather than nominal) GDP growth rates for the Western European countries for which the SNA was supposedly created (Arndt, Reference Arndt1987: 51).

These epistemological limitations were immediately understood by women statisticians and economists working for Britain’s Colonial Office, such as Phyllis Deane (Reference Deane1948), and UN officers at the DEA, namely Dudley Seers (Reference Seers1952; Toye, Reference Toye1987: 35–6). Amongst Stone’s assistants was Odd Aukrust, previously an aide to Ragnar Frisch. If Scandinavian work on national income acknowledged the importance of unpaid domestic work and financial assets, Aukrust recalled having no direct intellectual influence on the SNA’s formulation (Reference Aukrust1992: 47). The perceived shortcomings of the SNA’s Keynesian framework soon led the UN to engage with other accounting techniques. For example, the perceived need to detail inter-industrial exchanges led the UN Statistical Office to follow US practices and explore how to supplement national income accounts with input–output tables (UN, 1960a).

The search for alternative indicators further led the UN Statistical Commission to convene, in 1953, an expert group on the ‘International Definition and Measurements of Standards and Levels of Living’, which a decade later spawned the UN Research Institute for Social Development. Under the direction of Donald McGranahan and with the contributions of Nancy Baster, Cynthia Taft Morris, and Irma Adelman, the Institute conceptualised socio-economic indicators – from life expectancy to electricity consumption – as better gauges of wealth distribution than GNP/GDP (Baster, Reference Baster1972; McGranahan, Reference McGranahan1995). ECLA’s women thinkers, moreover, responded to the social and gendered inequalities they observed in fieldwork by developing their own methodologies to analyse the informal economy (Gautier-Morin, Reference Gautier-Morin2025). That work reflected the problems of equating ‘development’ with large capital investment projects, a conclusion that informed the 1970s ‘basic needs’ approach (Unger, Reference Unger2018: 133–42).

How then should we think of the UN System’s role in globalising national income accounting techniques that were contested even by its own economic experts? We can see that in post-war Western Europe, the SNA’s dissemination was a by-product of Cold War competition that attested to IGOs’s normative powers, particularly when aligned with US political aims. In helping to position reconstruction and development strategies around the pursuit of economic growth, the SNA also served to benchmark East–West economic performance (Schmelzer, Reference Schmelzer2016). Indeed, Mihály Simai’s international trajectory illustrates how IGOs acted as training sites for Socialist experts unfamiliar with Keynesian macroeconomics. The problem of finding equivalencies between the SNA and the Marxian-derived Soviet Material Product System generated decades of negotiations, involving the Council for Mutual Economic Assistance, the ECE, and the UN’s Statistical Office (Godard, Reference Godard2018).

We know that the SNA’s presumed ‘universal economic space’ (Schmelzer, Reference Schmelzer2016: 110) was always a functional fiction. Nevertheless, the UN’s global reach made it a powerfully performative concept. This was not least because national income accounting provided a statistical gridline for programming and measuring ‘development’, which became a defining UN objective since the 1950s (Toye, Reference Toye2006b: 829). In turn, growth measured in GDP became the threshold between ‘development’ and ‘underdevelopment’. That distinction was intrinsic to the UN’s First and Second Decades of Development (1961 and 1971), with their yearly GDP targets for North-to-South financial flows and South growth (Toye & Toye, Reference 84Turnell2004a: 176–9). UN technical assistance always involved strengthening national statistical expertise as a state-building function (Speich Chassé, Reference Chassé and Daniel2011), meaning that UN officers hailing from commodity-producing countries in the South were fundamental to the global dissemination of the SNA. This process involved the UN’s DEA and its Statistical Division, the UN Technical Assistance Administration (1950–8, eventually absorbed by the UN Development Programme), and the UN’s Regional Commissions. These organisations worked with non-governmental scientific bodies, such as the Inter–American Statistical Institute. Although aware of the need to improve sectoral statistical indicators, like agricultural output, the Commissions co-organised Conferences for ‘African’ and ‘Asian’ statisticians to discuss how to implement national income techniques in their regional contexts (UN, 1953; Serra, Reference Serra, Monsuru, Oladejo, Ajiola and Alabi2024).

The assumption that growth was determined by capital accumulation further explains the sustained attention given by the UN’s Statistical Division and the Regional Commissions, since the 1950s, to the difficulties of defining and measuring fundamental SNA aggregates, such as gross domestic capital formation, in ‘under-developed countries’ (UN, 1960b). ECLA’s structuralist support for infant industrialisation and regional market integration, following the Singer–Prebisch terms of trade thesis, was, accordingly, articulated through GDP growth rates, which, as discussed in Section 3.2, were forecasted using Keynesian models like Harrod–Domar. Although growth was an increasingly contested economic objective during the 1970s, GDP forecasts thus remained central to attempts at refocusing development policy on poverty, as well as to intergovernmental discussions about trade and monetary reform during the NIEO. The financial aid and foreign exchange gaps of commodity-producing countries facing declining or stagnant export prices were quantified in GDP (UNCTAD, 1968, 1973a). While acknowledging its inadequate rendering of social welfare, UNCTAD’s secretariat concluded that GDP growth rates remained ‘the best single indicator reflecting the combined effects of resource mobilisation and efficiency in their use’ (UNCTAD, 1973a: 16).

The omnipresence of GDP was similarly evident in the first mission of the ILO’s World Employment Programme, undertaken in 1971 by Singer to Kenya. Noteworthy for underlining the economic relevance of the informal sector, the defining concept resulting from Singer’s mission was ‘redistribution from growth’. Singer suggested that upper-class profits should be channelled into development programmes raising rural sector income, while assuming Kenya would grow at a sustained yearly rate of 7 per cent GDP (Shaw, Reference Shaw2002: 163–71). Notwithstanding its problematic reliance on the Keynesian growth theme, Singer’s report is considered to have foregrounded the ILO’s and the World Bank’s ‘basic needs’ approach to tackling poverty (Maul, Reference Maul2019: 171–81).

The 1973 oil crisis bolstered critiques of the faith in inexorable growth, but the result was not so much the ascendancy of alternative social indicators as the contestation of the UN’s monopoly in the production of GDP figures for the ‘developing world’ and, by extension, for the global economy. Despite the UN’s and the OECD’s statistical proselytising, the monetary comparability and publishing timeliness of national income accounts reinforced asymmetric understandings of the differences between the economies of wealthier and poorer nation-states (UN, 1968). This proved even more problematic after the Nixon Shock (1971), when floating exchange rates became the norm. Readapting OECD and UN statistics, the Bank and the IMF thus began compiling their own GDP figures to monitor governmental expenditure and debt. This process reinforced the importance of their own supervisory powers ahead of the 1980s Latin American debt crisis, which was another critical juncture in this global story of the influence and limits of Keynesian thinking (Ward, Reference Ward2004: 96–103). Such dynamics of competition and collaboration between IGOs in the international construction of macroeconomic knowledge encompassed both statistics and modelling.

3.2 Modelling Growth: ‘Harrod–Domar’, Two-Gap Models, and Project LINK

The aggregates comprising the SNA became integral to international economic thinking through macroeconomic forecasting and modelling. Modelling built on the UN’s statistical activities to support its ambitious policymaking research and state-building functions: from servicing national planning needs to reforming the Bretton Woods order (Toye & Toye, Reference 84Turnell2004a; Gay, Reference Gay2017). At the onset of the UN’s First Development Decade (1961), the operational importance of modelling was confirmed with the creation of an Economic Projections and Programming Centre (1962) within the UN’s DESA, as well as of Programming Centres in each Regional Commission (UN, 1962b). At stake was how the interaction between investment, savings, foreign aid, and trade could help define growth and development targets.

‘Development planning’ and ‘international trade and monetary reform’ were broad political and ideational platforms, and they were not only operationalised through Keynesian models. Since the late 1950s, the limitations of aggregative lenses and the failure of large capital projects in generating self-sustaining growth had prompted UN thinkers to explore other macro- and microeconomic techniques. Key UN collaborators such as Jan Tinbergen helped legitimise the UN’s technical assistance for development planning by advising on econometrics and linear programming (UN-ECAFE, 1960; Dekker, Reference Dekker2021: 287–308). This paralleled a gradual expansion in the UN’s statistical collections and in its member-states’s statistical capabilities. In 1968, interest in the inter-sectoral detail and disaggregated analysis offered by input–output programming led to its incorporation into the SNA’s first revision (UN, 1960a, 1962b; Ward, Reference Ward2004: 78). And in the 1970s World Bank and OECD, attempts to base project evaluation on a cost–benefit analysis that acknowledged social values drove new conceptualisations of ‘shadow prices’ (not to be confused with market-distorted prices) (Alacevich, Reference Alacevich2017: 274).

However, despite the diversity of modelling techniques employed by international secretariats, Keynesian toolkits had a lasting impact in the UN System – well beyond the 1950s heyday of Keynesian macroeconomics. In addition to the practical simplicity of those techniques (Toye, Reference Toye2009: 229), they held the appeal of mirroring a widely shared belief, across the UN’s economic secretariats, that GDP growth rates were primarily determined by capital accumulation. ‘Capital fundamentalism’ proved so intellectually momentous in the UN precisely because it was never exclusive to ‘Keynesianism’. To varying degrees, it underlaid most post-war development theories and overlapped with the political priorities of key postcolonial states (Toye, Reference Toye2009: 223). Against the perceived perils of a rising world population underemployed in the agricultural sector, the conclusion followed that raising living standards depended on improving labour productivity through capital investment (Boianovsky, Reference Boianovsky2018: 489). Thus, Arthur Lewis’s growth model with surplus labour was welcomed by Singer and his UN colleagues (Jolly, Reference Jolly2000: 92, 117; Shaw, Reference Shaw2002: 124). So too was the dual-sector model of Prasanta Chandra Mahalanobis, a member of the Indian Planning Commission and close collaborator of the UN Statistical Division (UN-ECAFE, 1960: 87; Ward, Reference Ward2004: 51, 165). Rosenstein-Rodan’s and Ragnar Nurkse’s theories of balanced growth, in which capital investment would raise demand and generate a multiplier effect between industrial and agricultural sectors, similarly became a staple of UN development discourse (UN, 1960c).

The quintessential macroeconomic growth model deployed in UN settings, however, was the Keynesian ‘Harrod–Domar’, which in 1962 spawned Hollis Chenery’s two-gap model. The following pages scrutinise their history, as well as that of Lawrence Klein’s Project LINK – the first macroeconometric model used by the UN to forecast short-term global growth, which was similarly based on a Keynesian articulation between investment, saving, and output. None of those toolkits was originally created within the UN System, but their secretariats proved central to their global dissemination and institutional survival.

‘Harrod–Domar’ is an ex post bracketing of Roy Harrod’s and Evsey Domar’s independently created theories, originally published in 1939 and 1948. Both drew on Keynes’s causal linkage between the multiplier and effective demand, and attempted to dynamise Keynesian theory by treating population, savings, and capital as variables that change over time. Initially concerned with understanding why industrialised economies hovered under full employment equilibrium, Harrod and Domar ultimately concluded that labour skill and technology determined growth in developing countries (Boianovsky, Reference Boianovsky2018: 479). Like most academic macroeconomists, UN thinkers instead assumed Harrod and Domar had unveiled the possibility of repurposing Keynes’s short-term theory of output into a long-term growth theory applicable to industrialised and industrialising countries. Janez Stanovnik, the Executive-Secretary of the ECE from 1968 to 1982, even attributed to Keynes the formula ‘that the national income increases with the increase of capital and the improvement of capital productivity’ (Weiss, Reference Wilkie2001: 50). The irony is that given conditions of inflation, undercapitalisation, and dollar shortage amongst commodity producers looking to industrialise, Harrod–Domar ranked domestic and foreign savings rates, ahead of insufficient effective demand, as the obstacle to development.

As shown by Boianovsky (Reference Boianovsky2018: 485–6), it was Singer who, in 1952, made ‘the first application of the Harrod–Domar equation to development planning’, thereby impacting the operational work of the UN’s DEA and the Regional Commissions. Singer’s equation associated economic development with income growth measured in GDP, as determined by the interaction between yearly population growth, saving rates, and capital productivity. By quantifying the (domestic) savings rate required to attain a desired growth rate in relation to a given capital-output ratio, Singer offered a practical way of quantifying the financial requirements of those UN member-states looking to industrialise. Given those member-states’s reliance on imported capital goods, that meant Singer’s equation backed UN attempts to boost North–South capital flows, with financial aid and import-substitution projects filling the income gap uncovered by international trade (Singer, Reference Singer1952: 11–9).

Through the 1950s, Singer’s take on Harrod–Domar reverberated across the UN’s Economic Commissions, as they added planning techniques to their technical assistance portfolio. Prebisch used Singer’s equation in 1954 to calculate target growth rates for Latin America (Boianovsky, Reference Boianovsky2018: 486), and Harrod–Domar permeated ECLA’s 1950s macroeconomic projections (UN-ECLA, 1955; Boianovsky, Reference Boianovsky2019: 158). Víctor Urquidi, a Mexican ECLA officer, recalled how the interest of César Furtado in Harrod–Domar (for which the domestic saving rate was a central variable) reflected the ECLA secretariat’s distrust of foreign direct investment as inimical to national sovereignty (Weiss, Reference Weiss2000c: 91; Boianovsky, Reference Boianovsky2010: 244–5).

In Bangkok, ECAFE’s Research Department immediately discussed Singer’s equation. Its staff acknowledged the dangers of taking capital-output ratios as stable and of seeing capital accumulation as a common development priority across Asia’s predominantly agricultural but highly diverse societies. The Research Department even recommended that, whenever possible, Harrod–Domar models should be complemented with less aggregative tools, such as input-output systems (UN-ECAFE, 1955: 39–40). However, ECAFE’s external collaborators, namely the Sri Lankan Gamani Corea (a future Secretary-General of UNCTAD), Tinbergen, and the Japanese economist Shinichi Ichimura, concluded that Harrod–Domar was appropriate to the poorest Asian nation-states because ‘capital is the most scarce factor in most under-developed economies’ (UN-ECAFE, 1960: 81).

The most politically visible application of Harrod–Domar by the UN secretariat occurred, nonetheless, in 1961. The DESA took advantage of US President John F. Kennedy’s proposal that the 1960s should be a ‘UN Development Decade’ as an opportunity to commit wealthier member-states to stable North–South aid transfers. This involved defining a scenario of yearly 5 per cent GDP growth rates in developing countries, a figure Singer and Prebisch arrived at after using Harrod–Domar (Singer, quoted by Toye & Toye, 2004a: 178–9). By then, Harrod–Domar was being superseded in academic debates by Robert Solow’s 1956 neoclassical growth model, which explained long-term growth through technological change, conceived as a variable exogenous to saving rates. Solow’s model deliberately ignored historical time and money (Halsmayer, Reference Halsmayer2024: 4, 10), two variables which in their statistical and policy forms were fundamental to the UN’s economic work. Indeed, Harrod-Domar’s aggregative perspective, like that of national accounts, was (paradoxically) less of a disadvantage for postcolonial countries struggling to build statistical expertise than the even more granular data required by input–output alternatives (UN-ECAFE, 1955: 40). Despite intra-UN critiques questioning the role of capital accumulation in raising the living standards of agricultural economies (Seers, Reference Seers1963), during the 1970s Harrod–Domar remained central to the UN’s macroeconomic projections, even for African countries lacking consistent statistical time-series (UN, 1972).

The long-term success of Harrod–Domar across the UN’s economic bodies rested on its simplicity and adaptability. Since the late 1950s, the model had been used by the World Bank to understand the long-term determinants of developing countries’s debt repayment prospects. Work conducted at the Bank by the Yugoslav economist Dragoslav Avramovič held that achieving long-term development goals required periods of indebtedness until growth through structural economic change could be achieved. Such change depended on the productivity of foreign capital inflows, which led Avramovič and his colleagues to analyse the interaction between domestic savings, investment, and capital efficiency (Laskaridis, Reference Laskaridis2021: 81–104).

Avramovič’s concern with the causal links between the foreign exchange gap, debt, and growth was shared by the Regional Commissions. Their operational experiences informed the emergence of two-gap models that complemented Harrod–Domar. ECLA’s attempts to formalise the macroeconomics of structuralism began with Harrod–Domar, but in 1956, the economist Osvaldo Sunkel reached similar conclusions to Rao’s (Reference Rao1952a) on the inapplicability of the Keynesian multiplier to commodity-exporting economies. Latin America’s reliance on imported capital goods, according to Sunkel, meant that Harrod–Domar’s reliance on domestic savings to generate self-sustainable growth was not an operative path for structural change. The need to acknowledge the trade-related foreign exchange gap thus informed ECLA’s empirical two-gap approach to calculating capital requirements (Boianovsky, Reference Boianovsky2010: 238–9, Reference Boianovsky2019). ECAFE economists were also aware of the importance of foreign exchange in development planning (Kitamura, Reference Kitamura1966), as were those of the OECD and the Bank, ahead of UNCTAD I (Laskaridis, Reference Laskaridis2021: 73–4).

It was, however, a former ECLA collaborator, Hollis Chenery, who, in 1962, while at the US Agency for International Development, first formalised two-gap modelling. Chenery’s two-gap model articulated Keynesian macroeconomic identities through linear programming and ‘opened up’ Harrod–Domar by prioritising foreign exchange inflows as a causal factor in the growth process (Boianovsky, Reference Boianovsky2010: 238–9). From 1971 to 1982, Chenery was the World Bank’s Chief Economist. During this time, two-gap models soon underpinned the Bank’s evaluation of financial aid requirements, at a moment when President Robert McNamara made poverty a prime issue for Bank lending (Easterly, Reference Easterly1999; Toye, Reference Toye2009: 229–30; Alacevich, Reference Alacevich2017: 279–80).

Two-gap modelling was also immediately adopted by the UN’s Economic Commissions (UN-ECAFE, 1968) and by UNCTAD. In 1965, ECAFE economist Hiroshi Kitamura (Reference Kitamura1966: 322–3) praised the utility of two-gap modelling while warning that overreliance on international capital assistance might prevent export-oriented economic reforms and create debt burdens that would thwart self-sustained growth. Launched in 1964 under Prebisch’s guidance (Toye & Toye, 2004a: 184–229), UNCTAD extensively used two-gap models. Although overlooked at the Bank, Avramovič’s findings were cited by Prebisch in UNCTAD I (Laskaridis, Reference Laskaridis2021: 75–80). UNCTAD employed two-gap models to formally legitimise its contested project of recalibrating the international division of labour in favour of commodity exporters. At stake was assessing their longer-term growth trajectories and quantifying their widening foreign exchange gaps (Behrman, Reference Behrman1971). Supervised by Sidney Dell and led by the Indian economist Vrudhula K. Sastry, a breakthrough result of UNCTAD’s two-gap modelling was the first collection of macroeconometric models for commodity-exporting countries from the Americas, Africa, and Asia (UNCTAD, 1968).

Sastry’s staff had benefitted from the input of Lawrence Klein and his closest disciples, especially Kanta Marwah, one of the first woman econometricians. UNCTAD’s objective of legitimising contentious policy proposals again explains its backing, in 1969, of Project LINK, an international modelling consortium launched by Klein that same year. Macroeconometric modelling was never an exclusively Keynesian endeavour, as evidenced by Tinbergen’s pioneering model of the US economy, built for the League’s Depression Delegation. Klein, however, was a leading figure in the mathematical formalisation of Keynes’s macroeconomics and the principal architect of post-war macroeconometrics (Pinzón-Fuchs, Reference Pinzón-Fuchs2009; Backhouse, Reference Backhouse2017: 516–31). After establishing his reputation with large-scale models of the US economy, which initially aimed to update Tinbergen’s model for post-war conditions, LINK can be seen as Klein’s attempt to ‘globalis[e] Keynesianism’ (Shenk, Reference Shenk2022: 1213). This meant broadening the macro-econometric lens beyond the single nation-state, albeit by bridging national and highly aggregated regional econometric models through their trade accounts (Klein, Reference Klein1976: 25).

The decentralised, cost-effective and collaborative modelling approach of LINK proved influential within the UN System and in supranational bodies like the European Commission (Rancan & Sergi, Reference Rancan and Sergi2024: 131–9). Because LINK held the promise of predicting global trade fluctuations and their impact upon commodity producers’s export revenue, the IMF (the prime provider of short-term financial assistance to the Global South) was amongst its early institutional supporters (Bodkin et al., Reference Bodkin, Klein and Marwah1991: 485–6; Blejer et al., Reference Blejer, Khan and Masson1995: 726–7). UNCTAD’s involvement soon extended to the UN’s Programming Centre and the Economic Commissions, which furnished LINK with the detailed statistical and econometric expertise that Klein required concerning both commodity-producing and Socialist states (Bodkin et al., Reference Bodkin, Klein and Marwah1991: 384–5, 395; Christian, Reference Christian, Christian, Kott and Matejka2018: 299–302).

LINK’s geographical coverage grew steadily, notwithstanding the inevitable epistemological problems related to its level of aggregation and the quality of the statistical data series underpinning its models. By 1979, its practitioners could claim to have created the first large-scale macroeconometric model of global trade (Bodkin et al., Reference Bodkin, Klein and Marwah1991: 490–6). If we look beyond LINK, following the Club of Rome’s 1972 report Limits to Growth, the 1970s witnessed alternative attempts at modelling planetary futures by forecasting demographic, technological, and environmental change (Andersson, Reference Andersson2018). An example was the UN’s Input–Output Global Model, directed by Leontief for the UN’s Centre for Programming. Yet by the early 1980s, the backlash against the NIEO meant that funding for large-scale modelling in IGOs largely dried up (Cole, Reference Cole1987: 404–16), whereas LINK’s short-term macroeconometric lens continued to underpin the GDP projections of the UN’s DESA.

3.3 International Monetary Keynesianism: Jacques Polak and ‘Mundell-Fleming’

Throughout its history, the IMF has been associated with the multilateral enforcement of conditionality through monetary and fiscal contraction (Sgard, Reference Sgard2023: 22–48). Compared to the belated ascendancy of economic expertise at the World Bank by the late 1960s (Hošman, Reference Hošman2023, Reference Hošman2024), economists always held equal status with lawyers at the IMF (De Vries, Reference Vries, Margaret and Coats1986: 53–66; Chwieroth, Reference Chwieroth2010: 40–60). The IMF recruited its staff heavily from the Bretton Woods delegations and Anglo–American universities. It even launched its own journal in 1950 to prevent the ‘premature burial [of its economists] in “an anonymous international organisation”’ (Polak, cited by Blejer et al., Reference Blejer, Khan and Masson1995: 707). Staff Papers enabled IMF economists (unlike their UN counterparts) to publish, under their own names, the results of confronting central problems of monetary policy, from exchange rate instability to whether monetary demand explained inflation (Polak, Reference Polak1996: 218). Leading the IMF’s Department of Research and Statistics from 1958 to 1979 was Jacques Polak, a former Tinbergen collaborator at the League who had joined the IMF in 1947 (Boughton, Reference Boughton2011). Polak (followed by John Marcus Fleming and Robert Mundell) made key contributions to international macroeconomics by stressing the importance of monetary variables in determining national income (James, Reference James1996: 82–3).

Interestingly for our purposes, Polak portrayed the IMF’s macroeconomic toolkits as Keynesian, even though they were designed to justify a type of conditionality that Keynes had critiqued. Certainly too, IMF historians reproduced the assumption that the Bretton Woods order and the IMF, specifically, were essentially Keynesian (Boughton, Reference Boughton2004: 13–4), and important early IMF recruits were indeed trained in Keynesian economics (Coats, Reference Coats1986: 135, 141). Margaret de Vries, one of the first (and few) women economists to join the IMF in 1948, was trained by Paul Samuelson at the Massachusetts Institute of Technology (Laskaridis, Reference Laskaridis2022). Amongst the IMF’s first Indian economists was, for example, Indraprasad Gordhanbhai Patel (I.G. Patel), a wartime student of Kahn and Kaldor (Berthelot, Reference Berthelot2001: 2–13).

Against that same background, it is worth exploring how IMF economists adapted Keynesian concepts and the IS–LM model of a closed economy to their operational mandate of sustaining international monetary stability. James Meade made the first significant attempt at a Keynesian open-macroeconomic model in Theory of International Economic Policy (Reference Keynes and Harrod1951). Meade was concerned with the relationship between domestic fiscal expansion and external balance under capital controls and fixed exchange rates, as stipulated in the Bretton Woods Agreements. His model, however, was too theoretical and dependent on detailed macroeconomic data to inform IMF policy work (Boughton, Reference Boughton2011: 384).

Meanwhile, in its early years, the IMF was struggling for international legitimacy. As Western Europe elided its monetary supervision due to the US’s decision to centre reconstruction efforts in the OEEC, the IMF secretariat began formulating its first conditionality criteria around the liquidity requirements of Central, South American, and Southeast Asian member-countries. Because of fluctuating export prices and crop seasonality, commodity producers were particularly reliant on the IMF’s short-term financial assistance. They were politically easier targets, too, for the Fund’s attempts to unravel the complicated post-war systems of multiple exchange rates used for protectionist ends and contrary to Bretton Woods (James, Reference James1996: 120–40; Babb, Reference Babb2007).

Indeed, the background to Polak’s Reference Polak1957 formalisation of the monetary approach to the balance-of-payments, which underlaid the IMF’s determination of lending criteria for the rest of the century (Boughton, Reference Boughton2011: 387), was a 1954–5 IMF mission to advise on the Mexican peso’s devaluation. Bank of Mexico officials sought IMF support for deflationary monetary reform (Thornton, Reference Thornton2021: 157–8). From their dialogue with Polak came the IMF’s monetarist formula that the money supply should vary with GNP (James, Reference James1996: 140; Polak, Reference Polak1996, 217). Polak’s model reflected the staff’s learnings from those first missions. It responded to a perceived need for a tool that legitimised the conditionality criteria specified in the IMF’s standby lending agreements, which included ceilings in bank credit and government expenditure (Boughton, Reference Boughton2011: 384; Sgard, Reference Sgard2023: 22–48).

From a theoretical viewpoint, what is striking in Polak’s model is less its rediscovery of how monetary factors could explain balance-of-payments disequilibria (which harked back to David Hume’s price-specie flow) than its reshuffling of Keynesian conceptualities to fit the quantity theory of money. This paralleled efforts by IMF economists like Sho-Chieh Tsiang (Reference Tsiang1956) to combine Keynes’s concept of liquidity preference (as the determinant of interest rates) with the neoclassical loanable funds theory. Polak was concerned with how the supply and demand for money would co-determine national income, that is, how a country’s domestic credit creation and trade-related external position determined whether income was spent domestically or abroad, and the consequent impact on the balance of payments. The General Theory’s multiplier forecasted how, for a closed economy, expansionary fiscal policy through the marginal propensity to consume would determine income. Instead, Polak eliminated the consumption function by assuming that ‘developing countries’ had low saving rates and low foreign exchange reserves. He related money demand (as determined by monetary policy and foreign exchange reserves) to income, an interaction animated by a multiplier where the velocity of money interacted with the ‘marginal propensity to import’ (Polak, Reference Polak1957: 1–19; Blejer et al., Reference Blejer, Khan and Masson1995: 712–6; Boughton, Reference Boughton2011: 384–5).

Not only did Polak radically reconfigure an axiomatic Keynesian concept (the multiplier) to define conditionality, but reflecting the IMF’s operational experience, his model was not entirely reliant on national income statistics. Given time lags in the production of national income data in Europe and Latin America and their questionable reliability (Boughton, 2001: 7), the IMF economist U Tun Wai (Reference Wai1957: 1–3), a former ECAFE official, explored how to calculate income from banking statistics. Using trade balances, money supply figures, and rough GNP averages, Polak’s model made it possible to survey IMF member countries regardless of the depth and time consistency of their national income accounts (Polak, Reference Polak1957: 3, 45; Boughton, Reference Boughton2011: 385–6).

The making of Polak’s model coincided with Per Jacobsson’s debut as IMF Managing Director. Jacobsson was a critic of Keynes who espoused a monetarist conception of inflation (Yago, Reference Yago2013: 92–6). During Jacobsson’s tenure, institutional agreement at the IMF that an export-led strategy was the optimal means to control inflation under growth (and without competitive devaluations) only grew. This was also the policy backbone of Polak’s model (Reference Polak1957: 19–37). Regardless, until the mid 1960s, the IMF secretariat identified itself as Keynesian because it stuck by the fixed exchange rates and capital controls mandated by Bretton Woods, which were amongst the few points on which Keynes and White had agreed (Chwieroth, Reference Chwieroth2010: 61–137). Yet, since the 1950s, the IMF was already pushing Southern American countries towards capital liberalisation (Calcagno, Reference Calcagno2021: 279–85). If IMF economists largely stood by capital fundamentalism as the vehicle to growth (Boianovsky, Reference Boianovsky2010: 241), backed by Polak’s model, they opposed ECLA’s favoured import-substitution as a cause of monetary inflation (Fajardo, Reference Fajardo2023).

When examining retrospectively his own work, Polak insisted that the IMF’s monetarist lens was not inimical to Keynesianism but rather a ‘evolutionary development of the ‘Kahn-Keynes multiplier’ (Polak, Reference Polak2001: 1), tailored for the statistics that the IMF staff could use for its operational needs (Coats, Reference Coats1986: 141; Emmerij, Reference Emmerij2000: 6). Despite their convergent conclusions over the negative impact of credit creation in foreign exchange reserves, Polak stressed that his tweaking of Keynes’s multiplier differentiated the IMF’s monetarism from that of the Chicago school, as formalised in the 1970s by Harry Johnson (Polak, Reference Polak2001). Ultimately, Polak’s reading of the General Theory was aligned with neoclassical critiques of Keynes and of what an optimal international monetary policy should entail. Polak thus characterised his Reference Polak1957 model as a historical example of the IMF’s ability to build a policy consensus that eventually informed the Washington Consensus (Polak, Reference Polak1996: 217).

If Polak’s model had a substantial policymaking impact (and evident socio-economic consequences), the Mundell–Fleming model of 1962–3 was groundbreaking in the theory of international macroeconomics. Robert Mundell joined the IMF in 1961 from the University of Chicago, where he had worked under John Fleming, then Vice-Director of the IMF’s Department of Research. Fleming had studied at the Geneva Graduate Institute and worked at the League’s EFO. Although he considered himself a ‘headstrong disciple’ of Keynes, they had disagreed in 1944 about whether post-war international commercial policy should be regulated by exchange depreciation, or by Keynes’s preferred quantitative restrictions (Fleming, Reference Fleming and Moggridge1944: 299).

‘Mundell–Fleming’ is an amalgamation of independent, if convergent, publications by those two economists. They expanded the ubiquitous IS–LM model into a short-term (taking prices and wages as fixed), two-country model (representing an open economy), which accounted for international capital flows and differentiated between the impact of monetary and fiscal policy on national income. Similarly to Polak, both were concerned with how money demand impacted the balance of payments, but they innovated by acknowledging scenarios of floating exchange rates and capital flows. In a fixed exchange rate system, monetary policy was seen as chiefly useful (compared with fiscal policy) to sustain balance-of-payments equilibrium. Instead, with floating rates and capital mobility, monetary policy would have stronger stabilising and expansionary effects than fiscal policy. Fleming’s and Mundell’s policy assumptions reflected the post-1950s return to monetary convertibility by the wealthier Western European countries and increasing market pressure for capital liberalisation. By showing how monetary and fiscal policy could determine income in different ways according to existing institutional monetary arrangements, their work foreshadowed the 1970s return to floating rates that allowed monetarist ideas to thrive (Boughton, Reference Boughton2002, Reference Boughton2004: 15).

Alongside the UN’s support of Project LINK, the impact of Mundell–Fleming demonstrates how the empirical work of intergovernmental agencies was intertwined with academic developments in theoretical economics (Backhouse, Reference Backhouse2024; Rancan & Sergi, Reference Rancan and Sergi2024). Exploring how IMF economists portrayed their monetary models as ‘Keynesian’ also allows us to see how the UN System’s economic work regularly intersected with normative discourses about the ‘meaning of Keynes’. This discursive dimension and its policy impact are nowhere more visible than in the successive attempts to reform the Bretton Woods order after the Havana Conference of 1947–8 and the ITO’s failure.

4.1 The Road to UNCTAD: Money, Commodities, and the Clearing Union

One early insight into the powerful symbolism of the late Keynes was how organised business responded to the 1948 ITO Havana Charter, when figures like the American businessman Philip Cortney and the early neoliberal Michael Heilperin succeeded in mobilising the ICC against the ITO (Slobodian, Reference Slobodian2018; David & Sampaio, Reference David and Sampaio2024). These early neoliberals portrayed both the Bretton Woods Agreements and the ITO as inflationary and protectionist, faults they attributed to Keynes’s nefarious intellectual legacy (Heilperin, Reference Heilperin1947: 88–100; Cortney, Reference Cortney1949). While Cortney and Heilperin denigrated Keynes, many former alumni of Keynes and Oxbridge were joining or collaborating with the UN. They included several Southeast Asian economists aware of the epistemological and policy limits of Keynesian macroeconomics when applied to undercapitalised commodity-exporting economies, including V.K.R.V. Rao. Yet they shared with their former professors – Nicholas Kaldor, Joan Robinson, Richard Kahn, James Meade, and Roy Harrod – a belief that international economic governance could be drastically improved through debtor-friendly, rules-based automatic mechanisms governed multilaterally. Here was how Keynes’s Clearing Union, perceived as a missed opportunity to build a fairer international order, started being repurposed to a decolonised world (Paesani & Rosselli, Reference Paesani and Rosselli2014: 4–5).

After the Havana Conference, commodity stabilisation debates converged with the problem of generating additional international liquidity to expand world trade and finance development (Toye & Toye, 2004a: 220–2), including Rao and Singer’s failed bid for a SUNFED. Reforming the international monetary system to fend off a possible dollar shortage during post-war reconstruction was a debate being stirred, in parallel, by the Belgian Robert Triffin, an OEEC and IMF economist from 1946 to 1951. Despite their divergent views on the quantity theory of money, Triffin keenly emphasised the broad similarities between Keynes’s Clearing Union and his own proposals to create a de facto international reserve currency, which would replace the gold–dollar Bretton Woods standard and be multilaterally managed (Maes & Pasotti, Reference Maes and Pasotti2021: 128–9).

One revealing example of how Keynesian networks became involved in international reform, before UNCTAD, was the FAO’s work on commodity price stabilisation. Before Havana, Sir John Orr, the FAO’s first Director-General, had attempted to launch a World Food Board (1946) comprising a buffer stock scheme to control the global production and distribution of basic foods. The Board became a dead letter after joint Anglo–American and IMF-Bank opposition. The same fate befell the FAO’s proposed International Commodity Clearing House (1949), which would have traded food surpluses for essential raw materials or channelled them into concessional food aid (Staples, Reference Staples2006: 82–104; Jachertz, Reference Jachertz, Plesch and Weiss2015). Heading the FAO’s Commodity Branch at its Economics Department during the 1950s was the German economist Gerda Blau, an LSE PhD close to Kaldor and Kahn (Rosselli, Reference Rosselli2017: 485–6). As discussed in Section 2.2, Kaldor was the first director of the ECE’s Research and Planning Division, from 1947 to 1949. Afterwards, Kahn spent 1955 assisting the Division with its yearly Economic Survey for Europe and advising on industrial policy (Marcuzzo, Reference Marcuzzo2020: 35).

When the Korean War (1950–1953) raised commodity prices and briefly made the stabilisation problem topical for commodity consumers, Blau commissioned studies from Kahn and from Kaldor, who would in 1962 also write a report on buffer stocks for Prebisch’s ECLA (Kaldor, Reference Kaldor1963). Kahn worked on an unpublished book manuscript with input from Joan Robinson and Meade. His proposed buffer stock scheme would have departed from Keynes’s by giving exceptional discretion to its administrators, empowering them to change price bands to beat market speculation. Yet like Keynes, Kahn aimed not at permanently raising commodity prices but, instead, compensating for market failure by upholding a political equilibrium between price stability, private profits, and productive efficiency (Toye & Toye, 2004a: 217; Fantacci et al., Reference Fantacci, Marcuzzo, Rosselli and Sanfilippo2012; Paesani & Rosselli, Reference Paesani and Rosselli2014; Rosselli, Reference Rosselli2017).

Before UNCTAD’s first session in 1964, and thanks to Prebisch’s advisor Sidney Dell, Kahn, Kaldor, and Meade were also involved in preliminary attempts to define common policy ground for expanding international liquidity and stabilising commodity prices. Kahn helped convince Dell to convene an expert group at Bellagio (Reference Kaldor1963) that brought together Blau, Kaldor and Pierre Uri, along with Oxbridge alumni who would later join UNCTAD, like I. G. Patel: a Kahn student and former IMF officer (Orange-Leroy, Reference Orange-Leroy2024). The Bellagio meeting originated a Kaldor proposal for a new composite international reserve currency named Bancor (as in Keynes’s Clearing Union), partly backed by gold and a commodity basket, intended to stimulate global economic growth (Paesani & Rosselli, Reference Paesani and Rosselli2014: 22–6). Co-signed by Tinbergen, Kaldor’s plan was submitted for discussion in UNCTAD I but was largely ignored. So too a paper by Meade that suggested avoiding buffer stocks altogether by using price-compensation agreements against falling commodity prices (Toye & Toye, 2004a: 214–7; Howson, Reference Howson2026: 604–8).

UNCTAD’s policy objectives comprised expanding world trade via preferential trade agreements (for the South’s exports), stabilising commodity prices, and creating supplementary liquidity for development finance. At UNCTAD’s Division for Trade Financing, Cambridge graduates from Southeast Asia, including Manmohan Singh and Lal Jayawardena – who had joined UNCTAD on Kaldor’s advice (Engerman, Reference Engerman2025: 164) – worked under Dell. Along with Gamani Corea, a Cambridge and Oxford graduate from Sri Lanka, Dell was also behind an UNCTAD I resolution to convene an expert group on international monetary reform. Its leading figures were Kahn and the Australian Trevor Swan, the co-creator of the Solow–Swan neoclassical growth model (Marcuzzo, Reference Marcuzzo2020: 34–6). The resulting UNCTAD report (International Monetary Issues and the Developing Countries, 1965) helped convince the IMF that commodity producers should have equal access to any newly created international monetary reserve and sparked debates over the latter’s linkage with development finance (Orange-Leroy, Reference Orange-Leroy2024: 11–27; Engerman, Reference Engerman2025: 164–7). On the sidelines, observers like David Carney (Reference Carney1967: 95), a Sierra Leonean economist at the UN African Institute for Economic Development and Planning, pressed the need for international monetary reform along the lines of Keynes’s Clearing Union. The outcome of those negotiations was the SDR: a fiat-based reserve asset, launched in 1969. Despite repeated comparisons with Keynes’s Bancor, SDRs were never automatically disbursed by the IMF, and the USA always rejected its ‘link’ with development finance (Wilkie, Reference Wilkie2011).

Although differences persisted between smaller and larger commodity producers in Latin America and Asia regarding the importance of export-led growth in development, UNCTAD reflected a degree of multilateral consensus about the need to reform the international trade and monetary systems. That consensus went beyond expert networks revolving around Keynes’s disciples, yet Keynesian toolkits (as seen in Section 3.2) helped define the economic knowledge sustaining UNCTAD’s policymaking, and discursive references to Keynes were used to justify the policies built upon that same knowledge.

This is apparent in the report Prebisch submitted to UNCTAD I on the trade gap between commodity consumers and producers, which was centred on GNP targets (UNCTAD, 1964: 3–6, 109). Dell provided Prebisch with a basic econometric model to back his case that to achieve the 5 per cent GNP growth target of the First Development Decade (which he had formulated with Singer), exports from developing countries should grow at a yearly 6 per cent rate, to avoid a trade gap relative to their imports of US$20 billion. (UNCTAD, 1964: 3–5; Toye & Toye, 2004a: 200). That epistemic reliance on aggregates later prompted former UNCTAD officers, including the Sri Lankan Lal Jayawardena, to (retrospectively) claim that Prebisch’s trade gap was compatible with Keynesian analysis, as the problem of development could be extrapolated from Keynes’s analysis of Britain’s full employment trade gap (Jayawardena, Reference Jayawardena, Haq, Mahbub, Richard, Paul and Haq1995: 52–3).

Throughout its foundational years, UNCTAD also mobilised the shadow of Keynes to legitimise its proposals vis-à-vis sympathetic interlocutors and against perceived resistance from other public and private international organisations. To Asian audiences, Prebisch presented UNCTAD’s efforts to create an international mechanism to finance buffer stocks as filling a gap left open by Bretton Woods, ‘notwithstanding the vision of Lord Keynes’ (Wilson, Reference Wilson1967: 19). Against business suspicion that UNCTAD wanted to interfere with market prices, Prebisch similarly told an ICC audience, in 1965, that linking commodity to monetary reform had been the objective of Keynes’s Clearing Union (Toye & Toye, 2004a: 222). When discussing with Prebisch, in 1966, about how to raise IMF and Bank support for UNCTAD’s reforms, one officer had no doubts that ‘Lord Keynes’s vision of the post-war order in trade and financial matters’, which included establishing an International Commodity Stabilisation Fund, remained ‘in the minds of both the leading people in the World Bank and IMF’ (UNCTAD, 1966: 5). Keynes was also retrieved by one of UNCTAD’s few women officers (and Marxian economists), the Indian Anila Graham, who had joined GATT in 1964 and edited UNCTAD’s Trade and Development Report from 1968 to 1976 (Sampaio, Reference Sampaio2023). As ‘Keynes so rightly pointed out’, Graham (Reference Graham1966: 6) concluded, expanding world trade depended on increasing international liquidity supply.

UNCTAD succeeded in mobilising other UN agencies to undertake work on commodity price stabilisation in the late 1960s, particularly the World Bank, until Robert McNamara’s tenure. Amongst the Bank’s key commodity experts was Dragoslav Avramovič (Hošman, Reference Hošman2023, Reference Hošman2024). He differentiated between the usefulness (which he questioned) of Keynesian demand analysis for explaining structural inflation in Latin America, and the relevance of Keynes’s support of commodity price stabilisation, in relation to UNCTAD-led efforts to tame cocoa price fluctuations (World Bank, 1966: 16, 39–40; Toye & Toye, 2004a: 219). When charting the theory of commodity markets (World Bank, 1968), Avramovič went back to Keynes’s Treatise on Money (Reference Keynes1930) and to a 1938 article which, unknowingly at the time, anticipated Keynes’s still ignored 1942 proposal for Commodity Control (Dimand & Dimand, Reference Dimand and Dimand1990) – I discuss in Section 4.2 the latter’s 1973 rediscovery by Lal Jayawardena.

Beliefs about the meaning of Keynesianism were similarly deployed, retrospectively, by former UN officers as an analytical framework for interpreting UNCTAD’s successes and failures during the 1960s–70s. Singer (Reference Singer1984: 14–15) characteristically narrowed UNCTAD’s diverse ideational background by tracing its historical origins to the ITO’s failure and Keynes’s support of buffer stocks. Interviewing former colleagues as a coordinator of the UN Intellectual History Project, Thomas Weiss, who was at UNCTAD from 1975 to 1985, queried his interviewees about whether the ‘UN and the people who worked for it are profoundly Keynesian’ (Weiss, Reference Weiss2000e: 49). Weiss conflated Keynesianism with public regulation of the market. The Zimbabwean Bernard Chidzero, head of UNCTAD’s Commodity Division from 1968 to 1977 and UNCTAD Deputy Secretary-General until 1980, concurred: UNCTAD’s core aim had been to ‘avoid [market] excesses or distortions’ (Weiss, Reference Weiss2000a: 62).

UNCTAD’s conception of multilateral economic governance was indeed anchored in expanding North–South interdependence by reforming capitalist trade relations. That third way, between unfettered market rule and an undefined global socialism, extended to UNCTAD’s institutional understanding of the 1970s NIEO. As industrialised countries and private interests contested the NIEO, UNCTAD’s staff continued presenting its policy agenda as following in Keynes’s footsteps.

4.2 From the NIEO to Brandt: Rediscovering Keynes, Redefining Keynesianism

The NIEO, launched by the homonymous UN Declaration of 1974, has been described as ‘a political brand holding together a set of loosely compatible agendas, which together formed something less than a coherent strategy’ (Gilman, Reference Gilman2015: 2). Although the NIEO mobilised significant political attention through the 1970s, the Declaration’s broad theme of strengthening national economic sovereignty indeed propelled intergovernmental and non-governmental bodies to pursue competing and often incompatible projects: from regulating multinationals and amending the inequalities of capitalist trade, to ideas of regional and global socialism. Despite that diversity, UNCTAD, as the de facto brains trust of the Group of Seventy-Seven (the main coalition of developing and commodity-producing countries in the UN), remained the prime intergovernmental agency tasked with giving the NIEO a tangible economic agenda (Rothstein, Reference Rothstein1979: 169–91).

UNCTAD’s NIEO involvement occurred not under the governance of the exhausted and disenchanted Prebisch, who quit UNCTAD in 1969, but rather under that of his successors, Manuel Pérez-Guerrero (1969–1974) and Gamani Corea (1974–1984). Corea wrote his doctoral thesis in 1952 on Sri Lanka’s export economy, under the supervision of John and Ursula Hicks. Amongst his conclusions was that stimulating effective demand via Keynesian expansionary policies would only aggravate the inflationary and structural imbalances characteristic of an undercapitalised commodity-exporting economy. Corea conceded, nevertheless, that ‘the instruments and techniques of analysis developed by Keynesian thought’ were helpful for understanding ‘the workings of an economy such as ours’ (Corea, Reference Corea1975: i–iii). With hindsight, Corea thought Cambridge had given him a ‘technocratic’ understanding of ‘the market system’, whereas Prebisch had forced him to ‘move away from the rigidity of orthodox formal economics’ (Weiss, Reference Weiss2000b: 9).

At UNCTAD, Corea’s main initiative was the Overall Integrated Programme for Commodities (1974), which was to be equipped with a $6 billion Common Fund to finance individual buffer stock systems. The Common Fund met with a slow death, after being accepted by a divided Group of Seventy-Seven at UNCTAD IV (1976) as a set of separate commodity agreements (Toye & Toye, 2004a: 210–53). Facing persistent suspicions from industrialised countries that commodity stabilisation was another means of transferring resources to the South through price fixing (Rothstein, Reference Rothstein1979: 60–3), the shadow of Keynes was evoked by UNCTAD’s staff throughout the NIEO. Despite awareness that Keynes’s aims for commodity stabilisation were more liberal than UNCTAD’s, its secretariat tried to draw political advantage from Lal Jayawardena’s belief that Keynes’s post-war reconstruction plans remained highly topical.

Close to Kaldor, Jayawardena was a Cambridge-trained economist, who, after working with Dell at UNCTAD, joined Corea in Sri Lanka’s Department of National Planning. Jayawardena was amongst the Southeast Asians informally bridging UNCTAD with the IMF, including as a delegate to the IMF’s Committee of Twenty, which began in 1972 as an attempt to democratise IMF governance. He lobbied for linking the SDR to development finance and for the IMF to finance buffer stocks, because of the falling price of tea, Sri Lanka’s staple export (Williamson, Reference Wilson1977: 94).

Drawing on his Cambridge background, in 1973–4 Jayawardena pitched to the Committee a paper detailing Keynes’s wartime support for commodity stabilisation (UNCTAD, 1973d; Helleiner, Reference Helleiner2014: 274; Engerman, Reference Engerman2025: 35–8, 272–3). Until then, Keynes’s views were known only from a brief mention in the 1943 Clearing Union proposals circulated before Bretton Woods. At the British Treasury archives, Jayawardena unearthed the earliest surviving draft of ‘The International Control of Raw Materials’ (Keynes, Reference Keynes1974). This was Keynes’s 1942 full-blown proposal for ‘Commodity Control’: a centralised buffer stock system dismissed after opposition ranging from the Ministry of Agriculture to the Bank of England (Keynes, Reference Keynes and Moggridge1980: 105, 112; Skidelsky, Reference Skidelsky2000: 234–6).

Jayawardena’s findings not only actualised historical knowledge about Keynes but also helped shape the discourse of international policymaking. What remains overlooked is that Jayawardena benefitted from UNCTAD’s guidance while researching and drafting his paper, immediately before the presentation in 1974 of UNCTAD’s Integrated Programme (Christian, Reference Christian, Christian, Kott and Matejka2018: 295). Advising Jayawardena was Alfred Maizels, the deputy head of UNCTAD’s Commodity Division and the intellectual driving force behind the Programme (Toye, Reference Toye2006d). Jayawardena (Reference Jayawardena, Nissanke and Hewitt1993: 10–2) found that Maizels had independently originated in 1968 (for UNCTAD II in New Delhi), the idea of a single Common Fund to administer a buffer stock network: the same idea Keynes pitched in 1942. Jayawardena’s enthusiasm contrasted with Maizels’s cool appraisal of how multilateral discussion of an updated Keynes plan could benefit UNCTAD. Maizels reasoned that if Jayawardena’s pitch to the IMF on a multi-commodity buffer stock system was successful, UNCTAD should conduct a ‘feasibility study […] as the agency responsible for international commodity policy’ (UNCTAD, 1973c).

Maizels praised Keynes’s proposal for a centralised buffer stock agency as a potentially cost-effective price-stabilising mechanism. Although the Integrated Programme revolved around a Common Fund, it was supposed to finance a series of individual buffer stock schemes, to be negotiated after separate commodity agreements were reached. Maizels, however, noted that, due to storage problems, buffer stocks ‘may not be as widely applicable’ to most commodities as Keynes had assumed. That, and the terms of trade problem, had prompted UNCTAD to explore alternative stabilisation policies, ranging from compensatory financing schemes to industrial diversification (Rothstein, Reference Rothstein1979: 81–102).

On those grounds, Maizels argued Keynes’s claim that stabilisation should not ‘change the underlying trend in the market’ was ‘extreme laissez-faire’. International commodity policy, Maizels continued, should use short-term schemes to improve long-term export earnings of commodity producers, ‘where practicable and also equitable for consumers’. His wording was a reminder of UNCTAD’s difficulties in defining what constituted fair remunerative prices (Toye & Toye, 2004a: 218–9, 243). Beyond its broader political topicality, for Maizels, the practical value of Keynes’s Plan resided in its detailed suggestions on export quotas and production controls. Maizels argued that because of the ‘current energy crisis’, commodity consuming countries might be interested in a ‘specially adapted’ Keynes Plan that would use its idea of sustaining production capacity ‘in excess of normal requirements’. Industrialised countries would be keen, Maizels reasoned, to preserve access to ‘adequate’ petroleum supplies (UNCTAD, 1973b).

Jayawardena ‘talked with Gamani [Corea] about the relevance of the Keynes material to UNCTAD’ (1973d) and presented his findings in a session of the UN General Assembly dedicated to the Integrated Programme. He even argued that Keynes’s pricing mechanism was more efficient than UNCTAD’s support of price indexation (UN, 1974; Rothstein, Reference Rothstein1979: 101). When advising on how international economic co-operation could be improved through UN institutional reform, Jayawardena cited UNCTAD’s Programme as ‘an adaptation of Keynes’s thinking to today’s circumstances’ (UN, 1975: 71). Through the late 1970s, UNCTAD staff, including Avramovič (Reference Avramovič1976: 1669), then on loan from the World Bank, continued to root the Programme in a historical genealogy dating back to Keynes’s Commodity Control. Maizels’s argument that the oil shock made Keynes topical came at a moment when political and intellectual support for fiscal fine-tuning and price controls was simultaneously being reinforced and undermined by monetary instability and stagflation. That explains how Jayawardena’s rediscovery of Keynes’s Commodity Control overlapped with the Johnsons’s coinage of the ‘shadow of Keynes’, as the North’s industrialised countries and business interests resisted the NIEO’s regulative agenda (Ogle, Reference Ogle2014).

A different response to the monetarist challenge against mainstream Keynesian macroeconomics was devised by the Economics and Statistics Directorate of the OECD, often described until the 1970s as the ‘house that Keynes built’ for its work on national income accounting and growth economics (Schmelzer, Reference Schmelzer2016: 191–2). As early as 1970, the OECD’s Economics Directorate began recalibrating its conception of macroeconomic demand management along anti-inflationary and market flexibility lines. Its key figure was the British economist Stephen Marris, a Cambridge PhD, albeit a mainstream Keynesian who stressed fiscal fine-tuning and embraced the Phillips Curve. After the oil shock, Marris was behind an OECD proposal (1975) suggesting that the USA, Germany, and Japan should act as ‘locomotives of growth’. Their domestic expansionism would be followed by monetary and supply-side retrenchment in other countries, supposedly creating a narrow margin for non-inflationary global growth. That idea was repurposed in the McCracken Report (1977), which reframed it in stronger pro-market terms. Before and after the second oil shock of 1979, the OECD began recommending supply-side ‘Positive Adjustment Policies’ as conduits of growth that were mutually exclusive from, rather than complementary to, boosting demand. After Marris left the secretariat in 1983, OECD research transitioned even more resolutely into supply-side views (Beroud, Reference Beroud, Leimgruber and Schmelzer2017; Gayon, Reference Gayon2022: 45–189).

Underlying Marris’s ‘locomotive theory’ was the postulate, dating from World War II debates, that North-led growth would trickle down to the South. The Brandt Commission’s Report, like UNCTAD beforehand, globalised that assumption by stressing the mutual benefits of expanding North–South economic interdependence through trade. It proved the last gasp of the NIEO ahead of the Latin American debt crisis, at a time when UNCTAD’s political and intellectual capital was fizzling out. Industrialised countries and leading commodity exporters rebuffed UNCTAD’s proposed Common Fund: there was no agreement on multilateralising price stabilisation for commodities like petroleum and coffee (Toye & Toye, 2004a: 245–53).

The Report originated with the Independent Commission on International Development Issues (1977–1983). Chaired by the former German Prime Minister Willy Brandt, it resulted from McNamara’s attempt to revitalise the debate on world poverty at the end of his term as World Bank President. Brandt operated independently from the Bank and the UN, but the secretariat drafting the Report under his direction had a strong international background. It covered opposite ends of the international economics spectrum, from Avramovič to the neoclassical Goran Öhlin, a former OECD economist (Stråth, Reference Stråth2023: 168–91).

The bestselling Brandt Report was immediately labelled ‘Keynesian’, even though Keynes was mentioned only when the Report praised the Bretton Woods Conference for enabling two decades of growing living standards worldwide (Brandt, Reference Brandt1980: 36). Its narrative revolved around economic interdependency. Aggregate demand from the North spurred growth in the South and vice versa, via world trade and North–South financial transfers. Although it qualified the pursuit of growth in environmentally friendly terms, it defined trade gaps, inequality levels, and growth targets in GNP/GDP. UNCTAD’s NIEO agenda also set the tone for Brandt’s proposals: regulating multinationals; opening North markets to South exports; and creating a Common Fund for commodity stabilisation. Echoing the 1950s SUNFED, the Report argued for substantial North–South financial transfers in the form of a ‘World Development Fund’, partly financed by an arms tax and structured as a central clearing house like Keynes had proposed (Stråth, Reference Stråth2023: 192–271). The Report’s globalisation of expansionary demand management was more than a response to the undermining of national economic boundaries by multinationals and capital liberalisation (Stråth, Reference Stråth2023: 52). It echoed a common theme of Keynes’s Clearing Union and Kaldor’s NIFE: macroeconomic management was indissociable from the multilateral coordination of monetary and trade relations.

The defining critique of the Report came from a dependency theorist, the Marxian economist Andre Gunder Frank (Stråth, Reference Stråth2023: 254–338). Frank slated the Report for its ‘Global Keynesianism’ and questioned the assumption that global economic challenges were reducible to sustaining demand. In Marxian lines, he depicted the world crisis as one of overcapitalisation rather than underconsumption – North–South dependency could only be broken if the South moved beyond capitalism through economic regionalisation. The Keynesianism of the Brandt Report was flawed, Frank continued, because it assumed that financial transfers would raise employment and living standards worldwide, while ignoring class conflict and power asymmetries within and beyond South nation-states. In fact, Keynesians such as Kaldor had long stressed the need to couple development strategies with the taxation of higher-income groups (Thirlwall, Reference Thirlwall1987: 130–58). For Frank, ‘Keynesianism’ in the North had only ever been possible at the cost of resource extraction in the South. Now, the Report sought to globalise Keynesianism when the North was already embracing monetarist retrenchment. Keynesian remedies, Frank concluded, offered no respite for stagflation nor from limited national autonomy in a capitalist world system (Frank, Reference Frank1980).

Frank’s notion of ‘Global Keynesianism’ has characterised historical depictions of the Report ever since (Stråth, Reference Stråth2023). Nevertheless, Singer and Jayawardena covered similar ground to arrive at different conclusions. They saw the Brandt Report as a pis aller that sidestepped the need for deeper structural reforms in national and international governance, ignored the Socialist East, and minimised oil producers’s unwillingness to redirect their credit surpluses into development finance. Unlike Frank, Singer and Jayawardena doubled down on the assumption that economic inequality could be tackled by an expansionary combination of market regulation and North–South financial transfers, which they framed as a progressive form of Keynesianism. Singer (Reference Singer1980, 1983: 197) called for a fairer ‘Keynesian International Economic Order’ that would draw directly from the Havana ITO Charter and Keynes’s Clearing Union, which he saw as a historical precedent to the NIEO. Shortly before assuming the directorship of the UN University World Institute for Development Economics Research (from 1985 to 1993), Jayawardena (Reference Jayawardena and Jansen1983: 149–74) similarly sketched an ‘International Keynesianism’ that would recycle the Organisation of the Petroleum Exporting Countries’s profits into Brandt’s World Development Fund.

Surveying how even development economists such as Singer were sceptical of the Brandt Report, Susan Strange (Reference Strange1981: 337), a pioneer of International Political Economy (Owens, Reference Owens2025: 265–90), noticed a growing incredulity about the possibility of rationalising politics through expertise, and by extension, about IGOs’s knowledge-based efficiency. Before the NIEO, Strange had warned that the UN’s and the South’s financial dependence on the West and capitalist markets would belie any attempts to radically transform the international economic system. The UN’s importance, she claimed, instead resided in its symbolic ability to persuade nation-states to shift their economic policymaking towards shared international goals (Strange, Reference Strange and Twitchett1971: 117–18). After the Brandt Report, Strange sensed that the North’s willingness to heed the technocratic advice of international organisations had waned for good (Strange, Reference Strange1981: 342).