Global economic order and global economic governance

Abstract

This issue of the Oxford Review of Economic Policy explores the origins of the US-led liberal multilateral economic order in the post-war world and the threats which that order now faces, drawing on contributions from two different groups of people—academic international relations (IR) scholars and international economists. This introductory essay attempts to weave the various strands of this intellectual collaboration together. First, it provides a narrative history of how economic collaboration emerged in the aftermath of the Second World War. Second, it describes the nature of the global economic governance that emerged and provides a new formal framework for analysing it, making use of the idea of ‘concerted unilateralism’. Third, it explores how contemporary challenges—a broadening of policy requirements, the rise of economic nationalism, and the rise of China as a new hegemon—mean that the global economic order is now in flux. And finally, it concludes with a general observation that runs through the paper: that IR scholars are inclined to analyse international economic regimes, economists to study particular policy proposals, and that these two perspectives can—and should—complement one another.

I. Introduction

In this issue of the Oxford Review of Economic Policy (OxREP), the contributors explore the origins of the US-led liberal multilateral economic order in the post-war world and the threats which that order now faces. The authors all ask—in different ways—how, in today’s challenging world, some form of cooperative international order might be sustained. In short, where might twenty-first century multilateral rule-making emerge from, and how might the resulting global economic order be sustained? These are important questions, given the obvious difficulties that this order now faces.

This introductory essay attempts to weave the various strands of this intellectual collaboration together. First, it provides a narrative history of how economic collaboration emerged in the aftermath of the Second World War. Second, it describes the nature of the global economic governance that emerged and provides a new formal framework for analysing it, making use of the idea of ‘concerted unilateralism’. Third, it explores how contemporary challenges—a broadening of policy requirements, the rise of economic nationalism, and the rise of China as a new hegemon—mean that the global economic order is now in flux. And finally, it concludes with a general observation that runs through the paper: that international relations (IR) scholars are inclined to analyse international economic regimes, economists to study particular policy proposals, and that these two perspectives can—and should—complement one another.

Our aspiration, as editors, has been to ensure that this issue of OxREP makes an original contribution to the discussion of these questions. To this end, we have brought together two different groups of people—academic IR scholars and international economists, (both academics and policy-makers). These people come from many different countries.

At the outset, we asked authors to consider how international cooperation in economic policy-making might best be achieved in the future. This has included the specific task of examining the challenges faced by global economic institutions, or GEIs (including the International Monetary Fund (IMF), the World Bank, and the World Trade Organization). But our collective activity has been broader than this.

This collaboration between IR scholars and international economists has brought together authors who bring different backgrounds, and methods, to their work. It has been our task, as editors, to manage this process of intellectual collaboration, and then to write about it in this essay. Doing that inevitably involves proceeding in a number of different directions at once. So this introductory essay is partly a piece of economic theory, but it also contains a number of policy proposals. It is also partly a paper about economic history, but it also contains material about the history of economic thought, in particular about open-economy macroeconomics and the theory of international trade.¹

Our endeavour has led us to appreciate the way in which papers by IR scholars can be used to frame our understanding of cooperation in the making of particular economic policies. That kind of cooperation is described in great detail—policy by policy—in the papers by the international economists in this issue of OxREP. The papers by IR scholars describe how these policies can be thought of as fitting together in an overall policy regime.

The first of the papers by IR scholars in this issue, that by Andrew Hurrell (2024, this issues) makes it clear that the complexity of the world economy is now forcing us to think again about the foundations of international cooperation in economic policy-making. We need to re-imagine—in a way which he describes—the nature of cooperation when there are not only international problems but truly global ones.

An overarching issue of concern, reflected in the title of this issue, has been whether global economic cooperation can—or should—continue to be underpinned by the values of liberal internationalism. These values have been important for the global economic order ever since the Second World War, even if they are now highly contested. The paper by John Ikenberry (2024, this issue) argues that the post-war liberal international order was successful precisely because liberal democracies, led by the US, had a particular set of capabilities which they put to good use in the support of a liberal international order. A companion paper by Jonathan Kirshner (2024, this issue) argues that these capabilities have rested on what he—and other political scientists—have called the ‘embedded liberalism’ that was present in the domestic political systems of those countries in the initial post-war period. As Kirshner and Ikenberry argue, this embedded liberalism is now under severe threat, throughout the world. Nevertheless, the paper by Dani Rodrik and Stephen Walt (2024, this issue) suggests how a kind of meta-regime for cooperation might still be built in such a world.

Our own paper proceeds as follows. In section II we provide a narrative history of how collaboration in relation to a fundamental set of economic policies was first established in the aftermath of the Second World War. We argue that this global order was successfully created only because there were significant changes in the economic policy-making framework throughout the world, set in train by national governments, managed by the IMF, the World Bank, and the General Agreement on Tariffs and Trade (GATT), and led by the United States of America as a new global hegemon. In what follows we use the term ‘Postwar International Policy Matrix’ (or PIPM) to describe the international economic policy-making regime that emerged. In brief this consisted of four things: the pursuit of full employment; the management of balance of payments adjustment by means of fixed but adjustable exchange rates; the regulation of international finance; and the promotion of trade liberalization. This section of the paper goes on to describe the way in which that PIPM was reconfigured after the collapse of the Bretton Woods system in 1971, and the creation of what became known as the global ‘non-system’. In that non-system, the pursuit of full employment was replaced by flexible inflation targeting, which tempers the pursuit of full employment with a concern for price stability, and the system of fixed-but-adjustable exchange rates gave way to a regime of floating-exchange rates.

Then, in sections III and IV, we deliberately step back to describe the nature of the global economic governance that emerged. Section III first sets out the conventional prisoner’s dilemma approach to the analysis of that governance that was developed, beginning in the 1970s, by both IR scholars and international economists. Then in section IV we put forward our new formal framework for analysing global economic governance, a framework which is rooted in the work of Young (1993, 1998a,b, 2024). We use the term ‘concerted unilateralism’, which we define, to help us think about that governance.

In section V, we use this idea of ‘concerted unilateralism’ to re-examine our historical account, bringing it forward to the present by describing the rise of economic nationalism and the problems which this is creating for economic policy-makers. Then, in section VI, we explore contemporary challenges to the PIPM and go on to show how policy requirements are now much broader than they were when the PIPM was first established, because of the need for global public goods related to climate-change policies and pandemic prevention. These challenges mean that the process of global economic governance is now in flux. In section VII, we look at the most important single shock to the global economic order that the world has seen since the 1940s—the rise of China as a potential new hegemon—and we spend some time considering how a system of concerted unilateralism might survive such a large shock.

Finally, in section VIII we draw some conclusions. We have already hinted as to what these might be, to do with how a set of policy proposals put forward by economists might be framed by the insights of international relations scholars. In our view, bringing these different groups of scholars together has yielded a significant interdisciplinary pay-off. We hope that readers of this OxREP issue will come to share this view.

II. The global economic order established after the Second World War

(i) The Postwar International Policy Matrix: establishing the ‘Golden Age’

In the second half of the twentieth century, a US-led liberal multilateral economic order emerged, one which was analysed in much detail at the time (see for, example, Schelling, 1960; Keohane and Nye, 1973; Waltz, 1979; Keohane, 1984; Kindelberger, 1988).

Laying the foundations of the liberal international order

In that global system, US hegemony filled the vacuum caused by the lack of leadership that had been evident during the 1930s. This position of leadership was thrust upon the United States by the Second World War. The Atlantic Charter, a joint declaration released on 14 August 1941, following a meeting between US President Franklin D. Roosevelt and British Prime Minister Winston Churchill, provided a broad statement of the ambitions shared by the US and the UK. In particular, both countries agreed to seek the liberalization of international trade, and to establish a set of labour, economic, and welfare standards. More generally, the Charter that Roosevelt and Churchill drafted included eight ‘common principles’ to which the US and UK would be committed in the post-war world.²

Negotiations between the US and the UK as to how these objectives might actually be achieved began immediately after that meeting between Roosevelt and Churchill, and were subsequently expanded to include other allied nations. Those discussions culminated in the Bretton Woods Conference of July 1944, at which the IMF and the World Bank were established, and in an international conference in Geneva in 1947, at which the GATT was signed. Importantly, what was created in those few years was a new international economic policy-making regime. That regime underpinned what the IR scholars John Ikenberry, Jonathan Kirshner, Andy Hurrell, and Helen Milner and Erik Voeten describe, writing in this OxREP issue, as the post-war ‘global liberal order’ or ‘liberal international order’ (see Ikenberry, 2024; Kirshner, 2024; Hurrell, 2024; Milner and Voeten, 2024).

That said, this use of the word ‘liberal’ is contested among IR scholars: from the standpoint of the Global South, for instance, one might sensibly ask whether it is right to call this order ‘liberal’ (see Oliver Stuenkel, 2024, this issue, and also Martin Guzman and Joseph Stiglitz, 2024, this issue). Nor was it the only order that prevailed during that period: until 1990, there was also the competing hegemonic system led by the Soviet Union.

In any event, the resulting international economic policy-making regime was managed by a set of policy-making institutions. Gone was the laissez-faire of economic policy-making which had bedevilled the 1920s and early 1930s.

In more detail, this regime had four components, each managed by an institution, or institutions, with its own particular task. First, national governments pursued Keynesian macroeconomic policies; there was a shared belief that full employment could be an objective of economic policy-making. Second, the International Monetary Fund (sometimes just called ‘the Fund’) supervised a system of pegged but adjustable exchange rates; there was an understanding that balance-of-payments imbalances between nations could be corrected by means of exchange rate changes, rather than by means of protectionism or austerity and deflation. Third, the International Bank for Reconstruction and Development (which became known as the World Bank, or just ‘the Bank’) facilitated movements of capital to the poorer parts of the world; there was a belief that international capital flows could be managed in such a way as to avoid international financial instability and, furthermore, that the international financial system could permit, and encourage, export-led growth by emerging market economies. Together the Fund and the Bank created an international financial framework which became known as the Bretton Woods system. (See the paper by Maurice Obstfeld (2024, this issue) for a detailed discussion of the history of that system.) Fourth, alongside this financial framework, the GATT was established as an interim regime—pending further negotiations on a proposed more comprehensive International Trade Organization—to facilitate the liberalization of international trade; there was a view that countries should grow and develop in an open manner, concentrating their economic activity on what their comparative advantage made it desirable for them to specialize in producing. In addition to these four components, this policy-making regime had a hegemonic leader, namely the US; and there was a belief that this leader could, and would, ensure that each of the above four components worked properly. In the present paper we call this four-part policy-making regime, along with the activities of its hegemon, the PIPM.

The Golden Age as the outcome of a negotiation

Of importance to this discussion is the fact that, in establishing this policy-making regime, the US sought to impose significant costs on the UK and its allies, in return for the financial support that the US would provide during the war.³ The passing of hegemonic leadership to the US, and the dismembering of the British Empire, was, according to the US, the price that the UK would need to pay for the defeat of the Axis powers. Article VII of the Lend-Lease Treaty, which the UK had been forced to sign at the beginning of the war,⁴ required that there would be provision ‘against discrimination in either United States of America or the United Kingdom against the importation of any product originating in the other country’. Specifically, this meant that the system of Imperial Preference would need to be abandoned.⁵ Crucially, Britain would also be denied the use of an Imperial Payments System (or indeed of any serious balance of payments restrictions) which this dismembering of its imperial markets might well make necessary.⁶

The resulting dispute between the US and the UK immediately led to a lengthy process of negotiation concerning post-war economic objectives and policies, led by Harry Dexter White for the US and John Maynard Keynes for the UK, a process which has been studied in great detail by both economists and IR scholars ever since.⁷ In relation to trade, the US wished to protect its producers in the US home market while insisting—at the same time—that countries within the British Empire not be able to grant protection to producers from elsewhere in the Empire in their own home markets. By contrast, Britain wished to achieve greater access to the US market for all goods from everywhere within its Empire. In relation to finance, the key question concerned how to ensure that there would be sufficient global liquidity to prevent the kind of collapse which had occurred in the 1930s and so enable the pursuit of full-employment policies by governments (the concern of Keynes) but not so much as to make an inflationary surge likely (the concern of Harry Dexter White).⁸

The hegemon’s role in influencing the agreement which was reached needs to be properly appreciated. Keynes was famously dismissive about US thinking when he said ‘They have the money, but we have the brains’, but in the end money won and White was able to push through the US conception of the IMF which was a watered down version of what Keynes had wanted. Nevertheless, for present purposes, the key point is that the US hegemon also did not get all that it wanted. The UK resisted—at least in part successfully—the attempt by the US to control the form of the UK’s post-war rehabilitation. Indeed, one can regard the agreements reached after many years—both at the Bretton Woods conference and in the establishment of the GATT—as embodying some kind of compromise between the US, the new global hegemon, and the UK, the hegemon which it was displacing. Although the US Congress sabotaged the establishment of the International Trade Organization, the establishment of the GATT ensured that the dismembering of imperial trade preferences, which the US sought, took place through a long-lasting and gradual series of trade negotiations in which the US—like all other nations—needed to make significant concessions. And although the pound was made convertible to dollars in July 1947 at the insistence of the US—causing an immediate monetary crisis—the institutions of financial cooperation which had been established at the IMF made the outcome more or less tolerable for the UK; there was an institutionally managed devaluation of sterling in 1949 and flows of international capital were strictly regulated.

This point is still of crucial relevance today, 75 years later, since, as we argue below, we will inevitably face future negotiations about the evolving world order, discussions which will be just as difficult now as they were then. It will be essential for countries other than the US hegemon to be able to make a difference in these negotiations,⁹ particularly given the contemporary rise of second-tier players with their own particular interests, including India (see Montek Ahluwalia and Utkarsh Patel, 2024, this issue, and Suman Bery, 2024, this issue); Singapore (see Shashi Jayakumar, 2024, this issue); Canada (see Jonathan Fried, 2024, this issue); Australia (see Gareth Evans, 2024, this issue); and the Global South (see Stuenkel, 2024), and given the rise of China (see Ross Garnaut, 2024, this issue, and Rana Mitter, 2024, this issue).

The post-war international economic policy-making regime which we have described helped to support material progress in what became a remarkably successful period.¹⁰ It became known as the ‘Golden Age’; and in France as ‘Les Trente Glorieuses’. The US, the global technological leader, enabled post-war reconstruction and then subsequent rapid growth in both Europe and Japan. And this spilled out into the rest of the world. It was a time of great progress, rather like the period before the First World War.

This regime was supported by domestic political conditions,¹¹ in both the US and elsewhere, described by political scientists as ‘embedded liberalism’ (see Kirshner, 2024). The essence of such conditions can be summarized like this: the community of those belonging to a particular nation felt connected to a common purpose, felt that the policies of that nation would protect them from any shocks which they might face, and that such policies would support them in a common collective project, in search of greater prosperity and well-being for all.

(ii) Reconfiguring the Postwar International Policy Matrix: the ‘Great Moderation’

The Golden Age was vulnerable since the first two parts of the PIPM—namely a commitment to full employment and a system of fixed exchange rates—unravelled in the late 1960s and early 1970s. What emerged was a new international economic policy-making regime very different from that put in place immediately after the Second World War; the components of the PIPM came to be deployed in a very different manner. Nevertheless, the result was, once again, remarkably successful.¹²

Bye-bye Bretton Woods: no more full employment policy or fixed exchange rates

By the mid-1970s, full employment Keynesianism had died as a system of domestic macroeconomic management. This was because there was a so-called ‘inflation bias’ in the system, a tendency to destabilizing inflationary pressures which occurred in different ways in different places.¹³ For whatever reason, confidence in the possibility of achieving full employment and stable prices began to fade away.

In turn, confidence also disappeared in the international aspects of the Bretton Woods financial system. International flows of privately owned financial capital increased significantly during the 1960s, and the pegged exchange-rates system came under increasingly severe speculative pressure. The devaluation of sterling in November 1967 was the beginning of the end. Four years later, the US government felt that it could no longer defend the convertibility of the dollar into gold for official holders, and in August 1971 it closed the gold window. The regime of fixed-but-adjustable exchange rates ended later that year after a conference convened at the Smithsonian Institution in Washington DC failed to keep the system together.

The result of these vulnerabilities was that the Bretton Woods system disintegrated as a system of global macroeconomic management.

Hello global ‘non-system’; inflation targeting and floating exchange rates

What took the place of the Bretton Woods system was a so-called ‘non-system’, characterized by the absence of international agreement with respect to the major features of the global economic order (Eichengreen, 2023):

The result was what we now call the non-system. There was no agreement following the collapse of Bretton Woods on a new exchange rate regime or system. There was no agreement on how to create a more symmetric adjustment mechanism. There was no global regulation of international liquidity. There was no enhanced role for the SDR [or special drawing rights].

And so, from the early 1970s a system of fixed-but-adjustable exchange rates gave way to this non-system of floating exchange rates which is now the standard for advanced economies.¹⁴

Such a set-up allows each individual country to choose its own monetary policy. After a confused period of nearly 20 years until the early 1990s—during which ideas about ‘monetarism’ led to chaotic experiments with monetary targets in a number of advanced economies—monetary policy came to be devoted to a policy of inflation targeting. Very remarkably, this is a policy set-up with a two-for-the-price-of-one feature, namely that one instrument—monetary policy—can be used to achieve two targets—both full employment and price stability—something which has been described by Olivier Blanchard as the ‘divine coincidence’ (Blanchard, 2006). In simple terms that idea goes like this. First, if inflation is equal to target, then monetary policy can, and should, be used to neutralize demand shocks so as to ensure that aggregate demand is kept as close as possible to aggregate supply, enabling resources to remain fully employed. But, second, if inflation is above target because of a supply shock, then monetary policy can, and should, be used to depress aggregate demand. This will bring inflation back to its target, at which point aggregate demand can again be brought into equality with aggregate supply and full employment can again be regained¹⁵ (see Allsopp and Vines, 2015).

In this set-up fiscal policy was—in the main—relegated to the pursuit of microeconomic objectives, subject only to the over-riding macroeconomic constraint that each country’s fiscal authority should remain fiscally solvent. See Kirsanova et al. (2009).¹⁶

This is a very different set-up from the Bretton Woods system. It took a full 20 years—until the early 1990s—for this non-system to fully emerge (see Stock and Watson, 2002), for reasons which Vines and Subacchi (2023) describe in some detail.¹⁷ And there were difficulties along the way including the disruptive aftermath of the East Asian Financial Crisis of 1997 which spread to other emerging-market economies, and the Mexican crisis of 1994 that had preceded it (see Fischer, 2005). What followed these events—which included the creation of the original Finance Ministers’ G20 and the implementation of a number of regulatory reforms—involved the creation of institutional structures that provided a more solid foundation for the global non-system.

The Great Moderation of the early 2000s

This new non-system, despite its name, was nevertheless a coherent regime. Indeed, it is possible to imagine how a perfectly functioning version of this new economic order—the ‘perfect model world’—might manage to replicate the international successes of the post-war Golden Age framework. Floating exchange rates could ensure external balance, against a domestic-macroeconomic-policy background in which each country pursued inflation targeting and maintained fiscal discipline. Such global macroeconomic coherence, could, at least in principle, leave the field open for a continuing liberalization of international trade, within a well-managed world economy, enabling countries to go on profiting from their comparative advantage as a means to grow as well as possible.¹⁸

In fact, this is what took place during the last decade of the twentieth century and the first decade of the twenty-first century. What was achieved during the period of the Great Moderation—Ben Bernanke’s term for the economic outcome during that period—was just as remarkable as what had been achieved during the earlier Golden Age in which the Bretton Woods system remained in place. Global growth was just as strong as in the Golden Age, and more broadly based.

However, this non-system now faces several contemporary vulnerabilities. Before exploring them, though, it is important to step back and ask a fundamental question. What was the nature of the global economic cooperation in the initial post-war period, and during the subsequent time in which the global non-system was put in place? How did these regimes come to be established?

We should make it clear, before proceeding, that we are not examining in detail the need for cooperation between countries at that time. This was, of course, due to the externalities, or spillovers, created by the policy choices which individual countries make. These spillovers are examined in great detail in the many papers of this OxREP issue. We are, instead, considering the form which that cooperation has actually taken. And in the later stages of this paper we consider the form which that cooperation will need to take in the future.

III. The initial study of international cooperation by IR scholars and economists

(i) The study by IR scholars of internationally cooperative regimes

The effectiveness of the post-war non-system, and the global economic cooperation that it secured, was a provocation for many in the field of IR: ‘the institutionalized policy coordination that we were increasingly observing in the late 1970s seemed an anomaly’, wrote the scholar Robert Keohane (2020). ‘How could it emerge consistent with the theory, or was something missing in the theory itself?’ At the time, the answer for many of these scholars was the latter. And they turned to ideas in economic thought, particularly game theory, for guidance.

Keohane, who would become the standard-bearer for this game-theoretic turn in IR theory, began with the work of Robert Axelrod (see Axelrod, 1984, and Axelrod and Hamilton, 1981). International cooperation could be thought of as a ‘prisoner’s dilemma’ game: though there are gains from multilateral cooperation, individual self-interested countries face a unilateral incentive to deviate and free ride. As a result, the cooperative equilibrium, where externalities are internalized leading to a Pareto improvement, instead gives way to the familiar sub-optimal Nash equilibrium. Axelrod’s contribution was to show that cooperation could still emerge among self-interested agents through reciprocity, under the ‘right’ conditions. In particular, the ‘shadow of the future’ might well prevent deviation and free-riding, simply because any game between countries is likely to be continually repeated. As a result, cooperation might appear to become beneficial because of the future benefits which a cooperative regime might bring. In important work together, Keohane and Axelrod applied these ideas to international politics, showing how these game theoretic insights from repeated games could be applied to make cooperation more likely (Axelrod and Keohane, 1985).¹⁹ International institutions were shown to have an important role in such a setting, by creating the ‘right’ conditions under which repeated games would lead to cooperation.

The other economic influence on Keohane’s thinking was the work of Ronald Coase, and in particular Coase (1960). In his eponymous theorem, Coase had identified three conditions under which unregulated bargaining, in the presence of externalities, might nevertheless lead to a Pareto efficient outcome (‘authoritative legal framework, perfect information, and zero transaction costs’). Yet what struck Keohane was not the impressive efficiency of the final result, but the stringency of the original assumptions: ‘it is absolutely clear that none of these conditions is met in world politics’ (Keohane, 1984). Again, this led Keohane to a similar realization as with Axelrod’s work: that a crucial function of international institutions was to create the conditions under which cooperation would be maintained. For that reason, this approach was known as a ‘functional theory of international regimes’ (Keohane, 1984).

There was another important strand to this IR work using game theory. A large literature emerged concerning the domestic political conditions for cooperation using the idea of so- called two-level games (see, for example, Putnam, 1988). The key, as emphasized in the paper by Jonathan Kirshner (2024), is that the outcomes in domestic political games influence the ability of countries to engage in, and sustain, international cooperation.

This work represented the start of a thriving field in IR, where scholars used game theoretic approaches to understand how international regimes of cooperation might emerge. For a review of this literature, see Zürn (1993). It continues to be influential today. Given that economists have historically made heavy use of game theory to understand international policy-making—an observation to which we return in a moment—this is an important historical detail.

This IR literature, which applies game theory to the problem of international cooperation, identifies theoretical conditions under which such a regime might be maintained. In doing so, it shows the difficulties of attaining international cooperation in the real world: the sorts of conditions in which free riding can be avoided are demanding and unlikely to be met in practice. For that reason, Keohane and others see an important role for institutions in creating those conditions.

One might then reasonably ask, how are those cooperation-sustaining institutions successfully built in the first place? Ex post, these institutions might help sustain cooperation among countries. But ex ante, how can those institutions be constructed without appealing to the same patterns of cooperation that they are themselves required to sustain?

This puzzle is not lost on IR scholars. ‘One major implication of the theory,’ writes Keohane, ‘is that the process of creating institutions is fundamentally different from the process of maintaining them, and much harder’ (Keohane, 2020). More generally, there is a recognition that game theoretic approaches to international cooperation raise more questions than they answer. What determines ‘the type of game that is played, in the first place?’ (Zürn, 1993); ‘What are the rules of the game? What are the choices available to each actor? What are the pay-offs in the game?’ (Snidal, 1985). These sorts of meta-question are often left unanswered. Yet in thinking about the emergence of international cooperation, understanding how answers to these meta-questions might emerge is extremely important.

(ii) The study by economists of international cooperation in the delivery of particular policies

International economists began to study aspects of this problem at much the same time. A conventional view of international macroeconomic cooperation, established in the economic literature in the 1980s, also suggested that any sort of cooperation ought to be hard to achieve.

An early focus in this literature was a macro-policy game being played by governments in the Bretton Woods regime (Hamada, 1974, 1979, 1985). The context was one in which fiscal policy was being used to manage aggregate demand and each country desired a trade surplus which—at the world level—added up to an outcome which would be impossible to achieve. The Nash equilibrium, without exchange rate movement, was a deflation war. Each country would pursue a trade surplus by means of fiscal deflation, at the expense of creating some unemployment. But symmetry between countries would mean that no country actually achieved a surplus; the deflation war would go on getting worse and worse until each country had so much unemployment that it was prepared to live without any of the trade surplus that it desired. A Pareto superior outcome in this simple set-up is one in which each country accepts that it cannot have a trade surplus and that, as a result, fiscal policy in each country is able to successfully pursue full employment. Of course, this analysis also pointed to the need to deploy two policy instruments (exchange-rate adjustment as well as fiscal policy) in the pursuit of two policy objectives (external balance as well as full employment) along with the need to establish a set of balance of payments targets which were globally consistent. The idea put forward by those who produced that early literature on international macroeconomic policy coordination was that the IMF should be thought of as an institution enabling exchange-rate movements, and agreements about balance of payments targets, of the necessary kind.²⁰

Another focus for international economists was the process of liberalizing trade. In theory, it was in the collective interest of countries to liberalize trade, in order to engage in some version of export-led growth and to specialize their production according to their comparative advantage. However individual countries, unilaterally pursuing their own narrow self-interest, might nevertheless resist the removal of protectionism, in particular the reduction of tariffs. In this framework, the liberalization that took place under GATT—a series of extremely difficult and slow ‘negotiating rounds’ of trade-liberalization negotiations, consisting of reciprocal reductions in tariff levels—was seen as a Pareto superior cooperative outcome, better than the protectionist Nash equilibrium. It was to be managed by the GATT, tariff reduction by tariff reduction, in a difficult manner.

A later focus in the macroeconomic literature was the prisoner’s-dilemma-like problem that emerged when the world moved to floating exchange rates. This was a time of high inflation. In theory, policy-makers faced a unilateral incentive to tighten monetary policy and appreciate their country’s exchange rate, in order to ‘export’ inflation to other countries. Successful cooperation would lead them to forgo such action. The idea put forward by those who produced that literature on international macroeconomic policy coordination was that the IMF should become an institution coordinating international macroeconomic cooperation of this kind (see, for instance, Oudiz and Sachs, 1984; McKibbin, 1988). But this did not happen.

There were other collective action problems that interested international economists during this period (e.g. the issue of finance during the Bretton Woods era, where the US and Keynes had clashed on the extent to which international financial flows ought to be restricted).

Taken together, the theoretical reasoning provided by this economics literature generated scepticism about the feasibility of multilateral international cooperation. And yet, in practice, despite these theoretical concerns, the Bretton Woods system and the subsequent global non-system were nevertheless policy-making regimes in which important cooperation still took place. Consider, for example, multilateral attempts to: restore growth in Europe after the collapse of the European Monetary System in 1992; respond to the East-Asian financial crisis of 1997–8; and contain the damage of the dot.com crash in 2002. And surely the best example of such coordination was the 2008 initiative of President Bush to convene the G20 at Leaders’ level in the wake of the Lehman-Brothers-triggered Global Financial Crisis together with the subsequent effort by British Prime Minister Gordon Brown in 2009 to use the G20 to put in place a coordinated response. In short, the theoretical scepticism about the feasibility of cooperation that is rooted in the prisoner’s-dilemma approach is hard to reconcile with the actual cooperation—albeit imperfect—that appeared to emerge in practice from the 1980s onwards. This suggests that the equilibrium concept, rooted in the prisoner’s-dilemma approach—a concept that was used by IR scholars as well as international economists, as we have just seen—is the wrong way of thinking about the challenges of international economic policy-making.

(iii) Taking stock: the need for a new approach

This brief survey suggests three things. First, economists have largely confined themselves to analyses of particular policies delivered one by one, and have attempted to understand how international cooperation arises in the delivery of such individual policies. By contrast, IR scholars have taken as their topic of analysis the construction of regimes which consist of the operation of a number of policies all at once. As we have seen, the post-war Bretton-Woods-plus-GATT set-up was one such regime, and so was the global-non-system-plus-WTO.

Second, it is clear that both approaches are important. A sustainable international economic regime requires cooperation among countries in the operation of each of a number of different policies. It also requires meta-level cooperation in holding the whole regime together. In particular, this will require that the pursuit of self-interested policies by a country or countries in any one particular policy sphere does not occur in such a way as to endanger cooperation in other different policy spheres. As and when that happens, the whole policy regime will come under pressure to unravel.

Nevertheless, third, it is clear to us that to properly understand the limited form of international cooperation in economic policy-making that has been sustained in the real world, one needs a new conception of what cooperative behaviour actually looks like, one that is different from the prisoner’s-dilemma scenario normally analysed by both economists and IR scholars. That conventional approach sees international cooperation as emerging when self-regarding Nash behaviour is somehow constrained by the ‘shadow of the future’ in the manner of Axelrod and Keohane. We believe that something rather different is needed.

IV. Towards an evolutionary approach to international cooperation

We now articulate what we think international cooperation in economic policy-making actually looks like. We go beyond the traditional analysis, adopted by both IR scholars and international economists, in three new ways. First, we abandon the prisoner’s-dilemma framework and replace it with a framework of ‘concerted unilateralism’ which leads to a game with multiple equilibria. We then allow for the existence of a hegemon. Finally—and crucially—we embed our new approach in an evolutionary framework. Each of these three things is necessary. Only by proceeding in this way can we arrive at a framework that is adequate for an understanding the past, present, and future of international cooperation in economic policy-making.

(i) Concerted unilateralism

As a first development, we think that international cooperation in economic policy-making actually involves a two-stage process.

In the first stage, global economic institutions (GEIs) are established, and the rules of the game associated with these institutions are determined. Then, in the second stage, countries take account of these rules when deciding what to do. They are free to choose what best suits them individually, within the framework that has been set in place by these rules.

We give the label ‘concerted unilateralism’ to this two-stage process.²¹ It is concerted because each country participates in the choice of the rules of the game, but it is also ‘unilateral’ because, once countries decide to play the game, they act individually, guided by their own self-interest, constrained only by the rules of the game that they are playing. Concerted unilateralism is, in other words, a process whereby states gather together to identify specific common objectives and constraints, but subsequently take steps to pursue their goals unilaterally and voluntarily.

The key idea of concerted unilateralism is best understood by means of two examples, one taken from policy-making in relation to international trade, and the other taken from macroeconomic policy-making. The first is drawn from the process of regional trade liberalization that took place in the Asia Pacific Region in the 1990s, under the umbrella of the Asia-Pacific Economic Cooperation process, or APEC.²² The second comes from the process by which policies of global fiscal expansion were engineered at the time, in 2014, when Australia was Chair of the G20.

In the Asia Pacific liberalization example, the story is as follows. When forming APEC, countries agreed to join in a process in which all were liberalizing their trade—the ‘concerted’ stage. Liberalization then became the strategy which all of the participants found to be in their own interest to pursue—the ‘unilateralism’ stage. And they each did this in their own way. Garnaut (1994a, pp. 280–1) explained how this came about.

Through the 1980s, and especially from 1985, there was an acceleration of unilateral, non-discriminatory trade liberalization in virtually all market oriented Western Pacific economies (including China). This led to observation that non-discriminatory trade liberalization was more likely in an economy that was part of a region in which other economies were liberalizing trade, than in a region of increasingly inward-looking states . . . In this world, each individual economy in a region [behaving in this way] observes the prosperity others are experiencing as a result of trade liberalization, in which the benefits of each country’s liberalization are extended by the contemporaneous liberalization of others. This increases incentives for each country to liberalize its own trade, bringing benefits not only to themselves but to others in the region. In this world, the regional trade policy ‘game’ takes [a new form which can] be contrasted with the ‘prisoner’s dilemma’ ‘game’ that has been used to define the process of trade liberalization amongst advanced industrialized economies in recent decades.

Figure 1 shows a hypothetical pay-off matrix for the trade-liberalization game being analysed by Garnaut. In each entry the first number denotes the pay-off to the home country and the second entry denotes the pay-off for a representative foreign country, supposed to be one of many other countries in the APEC region.²³

Before the APEC game begins, the pay-off matrix has this prisoner’s dilemma form for the following reason. If foreign countries remain protectionist, then the home country may well prefer to remain protectionist. See the footnote for an explanation of why this might be the case.²⁴ Protectionism at home will harm the foreign country, but by a small amount, because the home country is small. The same argument goes for a single representative foreign country; if the home country remains protectionist, then that single foreign country may well prefer to remain protectionist, and protectionism in that single foreign country will harm the home country, by a small amount, because the foreign country is also small. If both liberalize they each lose the large perceived benefits of protection but each gain a small amount because of liberalization in the other country.

The formation of the relevant GEI—here APEC—is supposed to change the rules of the game. A new setup is put in place in which all members of APEC can liberalize together, rather than each liberalising unilaterally, or just home and one foreign country liberalising at the same time. This is the ‘concerted’ stage of concerted liberalization. There are now increasing returns from liberalization; the more countries who liberalize, the more it becomes attractive to follow suit. In an earlier footnote we provided a quotation from the paper Axelrod and Keohane, published in 1985, which included the phrase ‘[i]nstitutions alter the pay-off structures facing actors,…’. That is what is happening here. Suppose, as a result, that the entries in the top left hand corner are now positive and greater than +1. Then there will now be two Nash equilibria, namely Protect-Protect (PP) and Liberalize-Liberalize (LL); Protectionism is no longer a dominant strategy.²⁵ This happens because the advantages to liberalising, when all countries within APEC are liberalising, are now much greater than if just two countries liberalize as a pair. See an additional footnote for an explanation of why this might be the case.²⁶ Of course for such a move to be beneficial in the way that we are here supposing, trade with other APEC countries will need to be large enough, relative to trade with the rest of the world, against which all members of APEC will now no longer be protectionist.

The change is concerted in the sense that everyone has agreed to be bound by the new rules, but within these rules players can still do as they wish unilaterally. Thus there is no assurance that protectionism will be abandoned immediately (it is still an equilibrium). Evolutionary game theory shows, however, that when the pay-offs from LL are sufficiently large (say at least 6), the PP equilibrium will tend to unravel under repeated plays of the game. This occurs because the (LL) equilibrium is now risk dominant: given complete uncertainty about the strategy of one’s opponent, it is safer to play L than to play P.²⁷ Under idiosyncratic shocks, the play of the game will tend to spontaneously move to the LL equilibrium and to stay there for long periods. The LL equilibrium is more stable than the PP equilibrium.²⁸

Note that if LL is the equilibrium which is chosen by each of our two countries—and by all of the other countries—they can each get on with the ‘unilateral stage’ of the concerted unilateralism process in their own way, each reducing the protection which it provides, on whatever products it chooses, at whatever speed it is capable of bringing about.

Gone is the idea that an agreement to cooperate is required in the everyday operation of the international system that the GEI oversees (here APEC). Once the rules of the game are established, the LL equilibrium will tend to emerge as each player acts in its own interest. As a result, we dispense with the idea that the ‘shadow of the future’ is necessary to ensure a cooperative outcome. Gone too is the kind of detailed surveillance and enforcement regime that is necessary to ensure the cooperative outcome in a prisoner’s dilemma game. The cooperative equilibrium LL emerges on its own, and once it is established, countries will not wish to free-ride in ways which undermine it.

Furthermore, the extract above from Garnaut (1994a) suggests that concerted unilateralism may be a more dynamic process than one which can be captured through the analysis of a simple one-shot game. Liberalization is not a discrete choice but a continuous process, where one country’s decision to liberalize further makes further liberalization more attractive to other countries, with the result that liberalization deepens among all countries in an iterative process of action which is mutually beneficial. Such a process is explored in much greater depth in Garnaut and Drysdale (1994). We turn to to consider the evolution of cooperation in this way in section IV(iii) below using the ideas of Peyton Young.

Of course, one must ask whether the rules of the game, and the LL equilibrium that results from these rules, are robust in the presence of shocks to the system. Countries may still wish to free-ride in ways which undermine the game, believing, perhaps falsely, that remaining protectionist is still advantageous as a strategy even if other countries liberalize. So, just as in a game of football, there are advantages in having a referee. That is because yellow cards—and the threat of red cards—may sometimes be necessary to ensure behaviour consistent with a successful game. In the absence of a collectively appointed referee, there needs to some other mechanism—in particular, a hegemon—to punish bad behaviour. We turn to the topic of hegemonic leadership in section IV(ii) below.

Such a trade-policy process was followed most closely in the western Pacific from roughly 1989 (when APEC was formed) to the time of the Asian Financial Crisis. During that period, from the mid-1980s onwards, countries in East Asia embarked on a process of concerted unilateral liberalization, which was described at the time as ‘Open Regionalism’. This process involved the most important liberalization of trade that the world has ever seen²⁹ with the partial exception of Chinese entry into the World Trade Organization in 2001. It ended when powerful US pressures for preferential deals³⁰ and the East Asian Financial Crisis combined to destabilize things and pave the way for the formation of preferential free trade areas.

In the example to do with coordination of global fiscal expansion in 2014, the story is as follows (see Vines, 2015, 2016; McKibbin and Vines, 2020). This was a time when Australia was Chair of the G20. In preliminary discussions at the beginning of the year, leaders from all nations were asked to agree that the G20’s focus for that year would be to raise the global growth rate, which was generally thought to have been inadequate in the 6 years since the Global Financial Crisis of 2008. It was also agreed that the method to be used to do this would be concerted fiscal expansion by all countries within the G20. Countries were then invited to propose policies which they might each embark upon on, so as to contribute to this overall global strategy. Each country’s proposals were reviewed by another country to see if what had been proposed was ‘fair’ in the sense of making an adequate contribution to the overall objective. It was agreed that, if necessary, countries would be invited to ‘do a bit more’ if the initial proposals were not large enough. This is what, in fact, was necessary to ensure that the whole set of proposals added up to a big enough total. So the process was, indeed, concerted.

For some countries—including the host country Australia—the process was a potentially difficult one, since the politicians in power were publicly committed to fiscal austerity, so that a commitment to fiscal expansion might well have gone against the announced policy intentions. However, countries were free to propose whatever kind of fiscal expansion they preferred. In particular, those countries committed to austerity were able to make proposals for increasing their expenditure on infrastructure, so that they would be able to display an asset which the fiscal expansion would create. It therefore became possible for them to avoid looking as if they were reneging on their commitment to austerity.³¹ So the process was indeed unilateral; it did allow for freedom of action.³² Indeed, Australian officials were at pains to ensure that there was no process of top-down instruction, and no top-down enforcement of conditionality.³³ It was clear at the time that the fiscal positions adopted by many such countries were more expansionary than they would have been had the G20 process not been put in place. A subsequent review judged the policy process to have been at least partially successful in influencing global macroeconomic outcomes.³⁴

Furthermore, as we will see below, this way of bringing about international cooperation is now being used—with some success—in relation to one other policy area, that to do with the reduction of carbon emission in relation to climate-change policy. And, as we will also see, this story about concerted unilateralism is an extremely simple way of explaining Peyton Young’s ideas—on which we focus below—about the emergence of cooperation as an evolutionary process. Those ideas of his are subtle and complex. But we believe that, in our simple exposition of the idea of concerted unilateralism, we have set out the essential contribution of those evolutionary ideas. The form of the pay-off matrix leads to a situation in which cooperation emerges from the bottom up and need not be enforced from the top down. Cooperation becomes self-fulfilling. This is the central feature of the ideas about the dynamic evolution of cooperation that we employ in our discussion in section IV(iii) below.

(ii) The importance of hegemonic leadership

A second development helps us to better understand the two-stage process which we have been describing. We view the post-war GEIs, and the cooperation that they sustained, as having had a pair of crucial features:

• (i) The everyday working of the post-war institutions resembled a ‘Stackelberg equilibrium’, led by the US hegemon. This introduces asymmetries of power: there is a ‘leader’ (or there are ‘leaders’) and ‘followers’. And there are asynchronies in action: in this two-stage game the leader acts first and then the followers respond.

• (ii) Those nations who joined in the Stackelberg game help to construct, and then subscribe to, the rules of the game. These rules were designed to prevent certain kinds of free riding, antisocial behaviour, not just by the Stackelberg followers, but by the leader himself.

In thinking about international cooperation—both in the making of trade policy and in macroeconomic policy-making—this approach captures three essential features of what happened. By making the US the ‘leader’, it captures one of the most distinctive properties of the post-war liberal multilateral order—US hegemony.³⁵ By introducing two stages, it makes clear that—in the first stage, when the rules of the game are chosen—the leader must take account of how the followers will respond to its decision-making. And third, by allowing for the freedom for each follower to choose its best response, after the collective choice has been made about the rules of the game, this approach recognizes that cooperation is limited in the second stage of the game. Nevertheless, that ‘freedom’ of the followers is conditioned by the actions and anticipated reactions of the Stackelberg leader.

(iii) Understanding the emergence of cooperation as an evolutionary process

A third additional development of our ideas seems essential. One major difficulty with our two-stage concerted-unilateralism story is that it is vague: the process through which the ‘rules of the game’ for international cooperation are agreed with the hegemon, the details of any rules that are agreed, and the way in which breakdown might occur in any arrangement is left unanalysed. And so, while this idea of concerted unilateralism, led by a hegemon, improves in terms of realism on the way of thinking about cooperation, as compared with a prisoner’s dilemma framework, it remains at best a metaphorical story of post-war international cooperation, rather than a tool for studying it.

We believe that the required elaboration can be provided by using the evolutionary analysis of cooperation put forward by Young (1993, 1998a,b, 2024). In this framework, equilibrium behaviours emerge from ‘the interactions and experiences of many dispersed individuals acting with fragmentary knowledge of the world, rather than from the actions of fully rational agents with commonly held beliefs as is often assumed in economic theory’. In practice, this means the models involve a large number of agents, each acting in a way based on beliefs formed from a snapshot of history of the behaviour of other agents, each willing to experiment and capable of making mistakes, all of whom gradually converge on some kind of equilibrium (Young, 2024, ch. 1). Young’s inspiration comes from the idea of spontaneous order, from the thought that social and economic structures can emerge from the bottom up, through the accumulation of many individual decisions, rather than being imposed top-down by a planner or authority.

In his forthcoming book, Peyton Young shows how contracts between a principal and an agent can evolve through a decentralized, adaptive process in which the expectations of agents are guided by precedent (Young, 2024, ch. 4). A given principal–agent contract specifies the ‘rules of the game’, that is, the terms that will govern their relationship. Once a set of rules are mutually agreed, the parties are constrained in how they can act. As play unfolds, a particular equilibrium will emerge, that is, the contract will generate pay-offs that are more or less desirable for the two parties as compared to other possible contracts. The choice of contract amounts to a meta-game that has the structure of a pure coordination game defined over the set of feasible contracts’ (Young, 2024, preface to ch. 4, and Young, 1998a). The choice of contract need not be once and for all, however, but is subject to revision as the protagonists learn from experience and sometimes experiment with new contract forms.

Young’s key finding is that this process leads to a situation in which one contract form is used by most people most of the time. Moreover, under certain circumstances, this contract is efficient in the sense that no other contract would yield strictly higher expected pay-offs for both parties. However, just as importantly, Young’s evolutionary framework also shows that it may take a long time before an outcome is realized between a principal and an agent which is equitable as well as efficient. As he says, societies can, and often do, become trapped in inequitable regimes and institutions that discriminate against some individuals or groups. Once such discriminatory institutions become entrenched, they can be very difficult to undo.

Young’s framework can be interpreted as a model of how institutions—construed as formal and informal rules of the game—evolve through the accumulation of precedent in the manner suggested by North (1991). Our idea in this paper is to apply his ideas to the formation of agreements not between a single principal and a single economic agent but instead to the agreements which have been reached by an international hegemon—i.e. the US—and countries in the international economic policy-making regime, in both the post-war policy-making regime and the global non-system which succeeded it.

Using Young’s language, we can add further structure to the idea of ‘concerted unilateralism’ that we introduced in section IV(i). Now, in the first stage, the ‘meta-game’, countries come together to choose a ‘contract’ with the hegemon that will govern their interactions. This is the ‘concerted’ moment, when countries make a collective choice about what game they are playing. Then, in the next stage, the ‘game’ proper, if a contract is agreed upon in the meta-game, countries interact according to the terms of that contract, but in ways that cannot be perfectly foreseen. This is the ‘unilateral’ moment, where countries act individually, in the manner of their own choosing, but constrained by the rules they had agreed on.

Young’s evolutionary model suggests two things, when applied to agreements reached by an international hegemon and countries in the international economic policy-making regime. On the one hand, unhelpful ‘rules of the game’ with inferior outcomes may stay in place for long periods of time. But on the other hand, more appropriate arrangements will eventually emerge, and—as Young shows—are inherently more stable once they become established. Of course, the time it takes for the long-run regime to be realized depends on a variety of factors. These include both the nature of the game and how complicated it is.

In addition, Young’s approach shows that conventions and norms which are established can come to have the property that, when the system is disturbed, it tends to return to the same position. Its ability to do this—its ‘resilience’—will depend, in part, on the size of the shock. It may be that the shock is small enough that the original ‘contract’ and the ‘rules of the game’ are both maintained.

Yet almost as important as how Young’s set-up can help us understand how the rules of the game might form and be maintained, is the fact that it can also help us think about how those rules might fall apart. If the shock is substantial enough, then the original rules of the game can collapse. New ones must be agreed upon to take their place. And this may take a great deal of time.

A very simple diagram of Young’s set-up is shown below: the US acts as principal, or leader, and the other countries are agents, or followers. The strategic interactions embodied in contract, or the rules of the game, operate in a simple back and forth manner between them, i.e. there are arrows pointing in both directions.³⁶

Young’s approach is at odds with traditional game theoretic approaches—in both the traditional IR and economic literatures. In those set-ups, the rules that govern how agents interact are either taken as given and the task is to study what equilibrium follows, or the efficient equilibrium is identified, and the task is to specify which rules will deliver that particular outcome (as in the mechanism design literature). Under an evolutionary approach, agents, full of behavioural imperfections, gradually work out rules of conduct and revise them in the light of experience.

It is important to note the importance of this last result about how cooperation can emerge and be sustained. In game theory, political scientists and economists have tried to secure a cooperative equilibrium in a variety of different ways, even though arguments about free-riding suggest that this will be difficult to do. In traditional one-shot prisoner’s dilemma games, for instance, they have used repetition of the one-shot game (e.g. Keohane and Axelrod’s early work, set out above) or made behavioural adjustments to agents’ preferences (e.g. introducing other-regarding or altruistic preferences). Young’s solution, though, is far more systematic: to replace the traditional game theoretic set-up with a dynamic evolutionary one in which cooperation—and the rules of the game that promote cooperation—can emerge dynamically without assuming that players are perfectly rational or have perfect foresight.

We think that this work by Young has significant implications for the IR discipline. In the 1980s, traditional IR scholars decided to take classical game theory seriously, and this decision revolutionized their field. A similar opportunity now presents itself to contemporary IR scholars; they could take seriously the evolutionary insights to be found in Young’s work. In what follows, we provide a sketch of how we think this might be done.

It is worth remarking that there already appear to be interesting similarities between Young’s economic formalism and the IR literature on ‘historical institutionalism’. In that literature, there is significant path dependency in institutions, whether or not they are fulfilling the valuable role that functionalists such as, for example, Keohane might imagine them to be performing (Rixen et al., 2016). Indeed, the opportunity which we see for the IR discipline might well involve IR scholars bolting Young’s economic formalism on to the already existing IR literature on historical institutionalism.

(iv) The evolution of international cooperation

It is revealing to use Young’s approach to the evolution of convention and norms to illuminate the history of the global economic order, its present state, and its likely future development.

To begin with, Young’s two-stage ‘meta-game’ and ‘game’ distinction captures what happened in the immediate post-war period. First, there was the formation of the rules of the game in the first-stage meta-game, which happened at the Bretton Woods meeting in 1944 which set up the IMF and the World Bank, and at the Geneva Conference of 1947 when the rules of the GATT were agreed. Together, these two dates were when the Postwar International Policy Matrix (PIPM) was initially established. Then, in the second stage, there was the play of the game which had been agreed upon, in the decades leading up to the 1970s, during which the Golden Age of post-war reconstruction and economic growth played itself out.

Then in the late 1960s and early 1970s things fell apart: wage inflation took off and movements of international financial capital led to speculative attacks on the fixed-exchange-rate system. These shocks were large enough to disrupt the prevailing rules of the game, and new ones had to be established.

Such a reconstruction actually happened. The PIPM was reconfigured, enabling the global non-system to emerge, with inflation targeting and floating exchange rates. There was then, subsequently, a play of the new game in the second stage. And so, in the early part of the first decade of the twenty-first century, during the age of the Great Moderation, the whole world grew rapidly, and much was achieved. We have noted that this re-forming of the global regime took a long time. The Great Moderation did not arrive until the early 2000s, a full 30 years after the collapse of the Bretton Woods regime. Such a delay is something which Young’s work would lead one to expect (see Young, 2024, ch. 1).

V. Economic nationalism

But the rules of the game are now disintegrating. This is much more serious than what happened in the early 1970s. In that earlier time period, the existing regime was threatened by inflation and by the collapse of the fixed exchange rate system. But the PIPM regime was re-configured. Now, it is the last component of the PIPM that is unravelling: the reliance on a continuing liberalization of international trade. This component of the PIPM had been supported until about a decade ago, underpinned by the view that countries should grow and develop in an open manner, concentrating their economic activity on what their comparative advantage made it desirable for them to specialize in producing. But the rise of economic nationalism is now challenging this view of the world.

As a result, we find ourselves moving away from the reconfigured PIPM. We noted above that Peyton Young’s approach shows that the conventions and norms of cooperation which have been established can be destabilized if the system is disturbed. If the shocks are small it will return to the same position, as happened after the increase in inflationary pressure and the growth of capital mobility which destroyed the Bretton Woods system. Although these shocks were large, the PIPM was reconfigured as the global non-system was put in place, culminating in the success of the ‘Great Moderation’ in the early years of this century. But, as we said, the world economy’s ability to do this—its ‘resilience’—will depend, in part, on the size of the shock. It may be that the shock is so large that the original ‘contract’ and the ‘rules of the game’ can no longer be maintained. New ones must be agreed upon to take their place. And this may take a great deal of time.

We think that this is what has now happened with the rise of economic nationalism, for reasons which we explore in detail in this section. We call the process of disintegration now taking place the ‘Great Unravelling’. As a result of this we are now again in a ‘pre-system’, like the world was in at the end of the Second World War; namely, we are back in the first stage of Young’s two-stage framework. The rise of economic nationalism means that the old rules of the game have melted away and the new ones will need to be put in place. In other words, our concerted unilateralism process is back in its first stage in which the process of concertation must again be undertaken.

(i) The challenge which economic nationalism presents for global policy-makers

In the last few years, the US and other major countries, including China and Europe, have moved in the direction of more nationalist, interventionist, and industrial-policy-oriented trade and technology strategies. The US has moved in this direction with its key trade and economic initiatives—the Chips Act, the Inflation Reduction Act (IRA), and its efforts to restrict export of high-end technology to China. Jake Sullivan laid out the US thinking in a widely discussed 27 April speech at the Brookings Institution (Sullivan, 2023). In effect, the Biden administration has indicated that the US position in the world economy is fundamentally changing.

For decades, the US’s leadership position was tied to its support for multilateral trade, reciprocal access to markets, and a continuous process of trade liberalization and expansion. This core pillar of the US position seems to be ending, or at least under radical reconstruction. We are entering into a period where US policy will be marked by more government intervention, industrial policy, export restrictions, technology competition, and sanctions.

It seems, to some analysts of US moves in this direction, that David Ricardo has not left the building, but that he is making room for Alexander Hamilton.³⁷

We now list three arguments which we have detected in the US in support of this economic nationalism. Clear thought shows that each of them presents significant risks to the prospects of global economic order.

(i) The first argument is about inequality. Globalization has made life difficult for low-wage, unskilled labour in the United States, and protectionism is said to be an appropriate response to this problem. Economists will recognize the Stolper–Samuelson theorem: ‘protectionism can increase the reward of the factor of production used intensively in the protected industry’ (Stolper and Samuelson, 1941). It is said that the political process in the United States does not provide any way of supporting workers who have been displaced by globalization other than through protectionism. See Kirshner (2024) in this issue.

Even though such a protectionist policy is seen by many as against the spirit of the WTO,³⁸ and said to be welfare-reducing by neoclassical economics, and probably supports low-tech, unproductive industries, it is politically necessary. This is the reasoning of Donald Trump’s supporters.

The support for protectionism for these reasons has resulted from the collapse in the ‘embedded liberalism’ discussed by political scientists and on which Kirshner (2024) focuses. The disappearance of this shared prosperity is not only important in itself. It has had the profound effect, as Kirshner has documented, of undermining the support in many countries for the liberal economic order.

(ii) The second argument is about innovation and is most clearly put forward by Martin Guzman and Joe Stiglitz in this issue of OxREP (Guzman and Stiglitz, 2024). There is a market failure to do with innovation. No innovating firm can fully capture the rewards which innovation creates, because the intellectual property is insecure since the newly invented ideas and processes are likely to be copied, and since its key workers are liable to leave, taking their ideas with them. Support of domestic innovating firms, and support of the domestic innovation process through the support of education and research is necessary, and if this is not possible, protectionism may again be chosen as a policy strategy. We call this response Stiglitz 1.

(iii) The third argument, which we call Stiglitz 2, is about dynamic comparative advantage (see Stiglitz, 1996). This idea can be understood by considering the policy experience of South Korea. In the early 1950s, neoclassical economists advised those making policy in Korea to formulate a growth policy based on static comparative advantage, which was to export primary commodities. The policy proposal was to develop the rice industry. This would have seen Korea locked into a development strategy with a very low growth potential. But Korea did the opposite by first protecting its industry from foreign competition and then opening to international trade, but only doing this when industry had been firmly established and was competitive enough to export profitably. This Stiglitz 2 argument is closely related to Stiglitz 1. It clearly does not only apply to Korea.

The question that arises for US strategic thinkers is: can the US remain at the centre of an open and rules-based coalition of states without continuing to underwrite an open trading system? But a deeper question also arises because of the response which other nations have been making by copying this policy of the US. Can the openness of the world economy be rendered stable when very many countries have become more government-driven, technology-competition-driven, industrial-policy-driven?

There is a point of principle at work here. Not only has US policy become protectionist. But it has done so in a selective manner. By contrast, right back at the beginning of the discussion with the UK in the early 1940s of post-war policy objectives, the US stood for the ‘most favoured nation’ (MFN) principle—and made this a core principle of the GATT. And it has remained a core principle of the WTO. This principle prohibits countries from discriminating between trading partners. It stands against the practice of making special deals for friends. It was that principle which enabled the US to stand so firmly against Imperial Preference within the British Empire.³⁹ Economic nationalism has thrown this important principle out of the window.

In short, economic nationalism is in danger of leading to an increasingly protectionist world. But the WTO rules about trade policy restraint now seem outdated and inappropriate as a response to that new struggle. Guzman and Stiglitz (2024) The original WTO rules were conceived using an economic theory based on competitive markets functioning under close to perfect competition. In such a world, protection would interfere with the workings of such markets, preventing countries from fully specializing in the production of goods in which they had a comparative advantage. Those like James Meade involved in the original setting up of the GATT in the mid-1940s, believed that trade rules were necessary in order to (gradually) remove this kind of protection through a series of international negotiations, so as to make international markets more competitive and thereby improve the efficiency of the world economy. But the modern world of oligopolistic rivalry presents a very different set of challenges. The old focus of the WTO on removing government-imposed protectionist restrictions of trade to allow markets to function more freely is not the right way to think about how the WTO should proceed in a world of oligopolistic rivalry. The paper by Guzman and Stiglitz (2024) discusses this issue in great detail.

Much of the new work that trade theorists will need to do about this topic has already been done by economists thinking about competition policy. Competition policy used to be based on old-fashioned neoclassical microeconomics, in which the appropriate action was thought to be the use of regulation to reduce monopoly power and make markets more competitive. But modern competition policy is much more subtle than this and sometimes relies on an argument that it may be good to subsidize particular monopolies, or allow them to set high prices, precisely because this might encourage endogenous learning.

For medium-sized countries—like Canada, Australia, and the UK—this new world creates two major kinds of challenges. The first is for each country to design its own industrial policies so that they do not degenerate into inefficient rent-seeking through protectionism, as in the past. The second is to find what role these countries can play on the international stage in writing the new global trade rules. If the US is in the process of ripping up the post-war trading rules, what can middle-level powers like, for instance, Canada and Australia do to help preserve the best of them? See Fried (2024) and Evans (2024).

Can a multilateral system of rules and institutions be devised to prevent the worst from happening? Where do China and the Global South fit into this shifting system? What coalitions of countries can work together to engage in rule-making in this area? Can the existing institutions be adapted for this purpose, or do we need new institutions? Can agreements be put in place to constrain abuses of sanctions and industrial policy so as to prevent a free-for-all?

The paper by Guzman and Stiglitz (2024) makes an initial and valuable step towards answering some of these questions. However, it seems to us that two major problems remain.

The initial problem will be to identify a dividing line between areas of international trade in which old ways of thinking remain appropriate, and areas of international trade in which new thinking is required. First, there are what we might call old-fashioned industries—like textiles, primary commodities and (perhaps) agriculture—where the commodity is relatively straightforward, and so where old ways of thinking remain entirely relevant. The task here is simply, as James Meade and his colleagues said in the late 1940s, to remove restrictions on trade, including tariffs, while at the same time—of course—avoiding environmental externalities and ensuring that the commodities are not made by slave labour or by children. This will remain familiar territory for the WTO. Then, second, there will be an intermediate category of goods and services which are much more difficult to think about, because they are produced with economies of scale, exhibit learning-by-doing and endogenous innovation, and are sold in oligopolistic markets. In the production of such commodities there is a first-mover advantage, so that revenues accrue to those who first begin production. They then have funds for further innovation, and can lower their costs and so race ahead. It will be the task of new policy rules to deal with the global regulation of markets in these products. Of course global policy-makers will need to identify which goods and services are like this, in such a way as to legitimize interventionist industrial policies in these industries but—at the same time—ensuring that such policies are only deployed in these industries. Having done this it will then be necessary to understand what policies of domestic support and subsidy for producers in these industries can be made acceptable in an international rules-based trading system. And it will be important to clarify how the resulting rules can be articulated and administered. Can this be done within the WTO? Or will new industry-based regulatory bodies be necessary? Finally, third, there will be goods and services which are strategically important for defence, such as chips and electronic circuitry, where restrictions on trade will be necessary and important. The defence-policy establishment of each and every country will have views about this and these will need to be considered. Of course, drawing a dividing line between these three categories of goods and services will be extremely difficult.

The subsequent problem is a global political-economy challenge of the first order. How might the resulting rules be administered in a fair manner, even although they will be intrinsically discriminatory. A beneficent global social planner would understand the difficulty here and would see that countries with industries which race ahead will need to reward—or at least compensate—those countries whose firms get left behind. Finding out just what these rewards should be, and who should pay them, will be a demanding task. The Guzman and Stiglitz paper provides important suggestions about how to analyse this problem of global social justice in Rawlsian manner. But none of this will be easy.

The 1930s loom as a warning. In a world that is actively engaged in restricting and sanctioning major sectors at the heart of trade and growth, protectionist and mercantilist spirals are only a few steps away. This is what happened with policies towards Japan adopted by both the British Empire and the US in that decade. But the early 1940s—and in particular the Atlantic Charter—also offer some promise. That Charter set out an international agreement on a set of objectives for the global post-war world, as we discussed in some detail at the beginning of this paper. The world now needs what would effectively be a new version of that charter. Such an agreement would show how economic nationalism could be put to one side so that—once again—countries could work towards a renewed common purpose.

(ii) We have been here before: the need to worry about more than the trade regime

Earlier, we noted that the US was determined to dismantle the preferential trading system which had been established within the British Empire in the 1930s. As a result, Keynes needed to ensure that the new rules governing the conduct of international trade would provide opportunities for Britain to prosper, and for the other members of the British Empire to do the same. The International Trade Organization, the initial rules for which were developed for Keynes by James Meade, and which laid the basis for what became the GATT—and ultimately the World Trade Organization—were to regulate trade in a multilateral, not unilateral, manner. As a result, Britain and the countries in its Empire would therefore not be forced to liberalize their borders unless the US itself liberalized, and unless a global system were established in which the liberalization process would be managed in a cooperative multilateral manner.

Nevertheless, as the Second World War drew to a close, many in the UK, and elsewhere, were concerned about the threat that the UK faced from the US in relation to this matter. They thus questioned the strategy which Keynes had adopted in the run-up to the Bretton Woods Conference in July 1944. They asked him why he had devoted so much attention to international macroeconomic management, and to the establishment of the IMF and the World Bank. Why had he not, instead, gone immediately to a discussion of the key issue: how to manage international trading conditions. Why had he not immediately moved to establish an International Trade Organization, rather than allowing those negotiations to be postponed and not completed until they were finalized at a conference meeting in Geneva in 1947, three whole years later?

One of us has already discussed this matter in print in some detail (see Vines, 2017, 2023). But it is worth quoting here Keynes’s response to that question, since it is of relevance to what we say in this paper. In a speech in the House of Lords, just before the Bretton Woods Conference was about to begin, Keynes said:

There was a logical reason for dealing with the monetary proposals first. It is extraordinarily difficult to frame any proposals about tariffs if countries are free to alter the value of their currencies without agreement and at short notice. Tariffs and currency depreciations are in many cases alternatives. Without currency arrangements you have no ground on which to discuss tariffs. . . . It is very difficult while you have monetary chaos to have order of any kind in other directions . . . if we have a firm ground on this particular issue it will be a great deal easier to reach a satisfactory answer on other questions. It is perhaps an accident that the monetary proposals got started first . . . but I am not sure that it was not a fortunate accident. (Keynes, 1944)

It was this conviction which led Keynes to proceed in the way in which he did. It is as a result of a similar conviction that we believe that it is essential for global policy-makers to concentrate not just on the future of global trading relations but also on the other aspects of the PIPM. And as we will see there is even more to do than this, given the need to provide global public goods of a kind which were not seen to be important back in the mid-1940s.

VI. Finding new rules for the global-economic-policy-making game

We said earlier that our process of concerted unilateralism must go back to the first stage in which new rules of the global policy must again be put in place. What problems will these new rules of the game need to deal with?

(i) Finding a new version of the PIPM

As we have described above, the PIPM had four components, each managed by an institution, or institutions, with its own particular task. National governments pursued Keynesian macroeconomic policies; the International Monetary Fund supervised a system of pegged but adjustable exchange rates so that balance-of-payments imbalances between nations could be corrected by means of exchange rate changes; the World Bank facilitated movements of capital to the poorer parts of the world; and the GATT (which became the WTO) was established to facilitate the liberalization of international trade. In addition, this policy-making regime had a hegemonic leader, namely the US.

We described the challenges to the role of the WTO in the previous section. We discuss the challenges to the US as hegemonic leader in the next section. But there have also been challenges which have made it difficult for national governments, the IMF, and the World Bank to fulfil their roles. We now describe these briefly.

Issues for national governments: the need for more than Keynesian macroeconomic policies

There has been a loss of confidence in the idea that inflation targeting should be the central focus of domestic macroeconomic management—even if it is coupled with a concern for maintaining a high level of employment. There are many reasons for this. Many industries in the modern world are not competitive, because the high fixed costs and low marginal costs make imperfectly competitive outcomes inevitable. So, there is a story to tell about the huge returns in FANG markets (Facebook, Amazon, Netflix, Google), both for those who own the assets and for those who work in these industries. This is something which has fed a massive rise in real incomes at the upper end of the income distribution. It has been coupled with a very large sectoral divergence in productivity of modern parts of the economy (mainly in the tradeable goods sector) from the more traditional parts (mainly to do with the production of non-traded goods and services). That has led to a massive fall in the relative rewards of those at the bottom end of the income distribution. That has been coupled with the effects of globalization on trade which have severely damaged the real rewards of many of those in sectors subject to import competition. Substantial immigration, including of relatively low-skilled people, has added to these pressures.

Issue for the IMF and the World Bank: financial instability

Turbulence in the global financial system led to the global financial crisis (GFC). The Great Moderation—that period at the beginning of this century which looked to all intents and purposes like a perfect model world—came to an abrupt end in 2007. The financial collapse, and subsequent deleveraging, led to the most rapid decline in economic activity since the Great Depression.

The initial policy response to the crisis by world leaders displayed remarkable international cooperation. At the G20 Summit in London in April 2009, global policy-makers worked in close cooperation with the IMF and the expansionary macroeconomic policies agreed at the time—both fiscal and monetary—played a significant part in preventing the GFC from causing a fall in output comparable to that which happened during the Great Depression.

Nevertheless, 10 years on from that time, the brutally oligopolistic, and short-term, nature of the global financial system continues to worry policy-makers.

It is clear that there have been large improvements in financial regulation in advanced countries. And it is true that the onset of Covid did not lead to massive financial collapse, which it might well have done. But it is still not clear that the regulations have been adequate (see Obstfeld, 2024).

Furthermore, developing countries and emerging market economies continue to face risk-on risk-off pressures coming from global financial markets, leading to speculative inflows and outflows of financial capital. This has significant policy implications for the IMF which Stephen Grenville, Dede Basri, and Jonathan Ostry discussed in the recent issue of OxREP on international monetary reform (see Grenville, 2023; Basri and Sumartono, 2023; and Ostry, 2023). As yet, the IMF does not seem sufficiently willing to use the necessary policy tools at its disposal nor to encourage countries to use capital flow management where this seems necessary.⁴⁰

As Ngaire Woods (2024, this issue) notes, other global challenges—debt distress, post-Covid development crisis, unwinding of quantitative easing and unconventional monetary policy—are creating additional demands on the IMF. But in each case, the IMF is hamstrung: on the one hand, by the rise of economic nationalism which undermines the pursuit of globalization as a core IMF principle, and on the other by the strengthening of geopolitical rivalries which make members less willing to cooperate on any particular problem. (See also Obstfeld (2024) for an account of the way in which what he calls the slowdown in globalization is magnifying the policy problems which need to be dealt with.)

Woods argues that the larger are the divergences between the values and purposes of the principal shareholders of the Fund, the more difficult it will be to devise rules of conduct for the executive staff members of that institution to follow in performing their duties, and the less likely it is that the institution will be able to fulfil its tasks and purposes. It is valuable to have the implications of this idea spelled out in detail for a particular institution. This tension clearly also challenges the WTO and the World Bank.

(ii) Going beyond the PIPM

Issues for the WTO, the UN, and the Word Bank: managing the transition to net-zero carbon emissions

Climate risk has become a major policy concern in all countries. This is a global problem, and the solution will necessarily involve global cooperation. That cooperation is being managed by the Conference of Parties (COP), the decision-making body of the UN Framework Convention on Climate Change (UNFCCC). It will need to fly in the face of the strategic competition described before.

The rough combination of action that will be required is now familiar. This includes, but is not limited to, a collection of well-known mitigation and adaption measures: increased carbon taxes; decreased fossil fuel subsidies; improvements in carbon storage; de-carbonized energy generation; adoption of sector-specific low-carbon technologies; and many others. To see these interventions in China and India, consider Garnaut (2024) and Ahluwalia and Patel (2024), respectively.

The inadequacy of climate-change-related policies which the world has undertaken is well understood. In particular, the inadequacies of the Paris Agreement in 2016 did not take long to surface—even though the Paris conference was declared to be a resounding success. The Inter-governmental Panel on Climate Change (IPCC), in a report published in 2018, warned that the remaining carbon budget for the world, consistent with limiting global warming to +1.5°C, was shrinking and the world as a whole had to reduce emissions to net zero by 2050 to keep global warming within the limit set. The adequacy of the Copenhagen commitments always depended on the ratcheting up over time that was anticipated within the Paris Agreement. Despite subsequent progress in Glasgow, the sum of the commitments remains much weaker than is required to achieve agreed global climate goals.

Nevertheless an approach along the lines of concerted liberalization is now yielding fruit. This new approach, foreshadowed in the discussions, late in the Copenhagen conference, between President Obama and the representatives of Brazil, China, India, and South Africa was given form at Cancun a year later, and more concrete form at Paris and Glasgow. The Cancun/Paris approach had countries putting unilateral voluntary proposals forward. They would be discussed by peers and amended every 5 years so that a set of commitments that did not add up to 1.5°C at Paris would be gradually upgraded so that, if all worked well, they would eventually be strong enough to meet the global goal. So the Paris commitments (with India’s particularly weak) did not add up to 1.5°C. But the Glasgow revisions came closer (with India’s ‘net zero by 2070’ providing a commitment to net zero for the first time). Australia had been the only developed country without a ‘net zero by 2050’ commitment before Glasgow, but corrected this at Glasgow. The world is still not there for 1.5°C or even 2°C. But it is coming closer.

As Montek Ahluwalia has said to us, one can think of ‘concerted unilateralism’ as a second-best kind of decision-making process. In his paper with Utkarsh Patel, he makes the point that, while the logical way to proceed in setting decarbonization targets would be to agree on a rational way of dividing the remaining carbon budget for the world across major countries, this has not been attempted in COP meetings because it is realized the countries would not be able agree on a formula.

So the alternative was to keep nudging the herd to a ‘more’ common point of view. The whole story of developing countries in climate change negotiations reflects [the relevance of the unilateralism part of the concerted-unilateralism process. These countries] first refused to accept any obligation, then agreed to reduce ‘emissions intensity’ and finally agreed to actually reduce absolute emissions and get to net zero albeit at different dates. This is a perfect example of [the concerted-unilateral process at work]. Of course it takes a long time.⁴¹

It is in this process of continuing refinement of unilateral voluntary proposals that progress seems to be taking place towards the desired objectives. Gone are the targets and timetables of the Kyoto approach, which were based on a prisoner’s dilemma approach to international cooperation about climate policy. As Garnaut (2024) also describes in this issue of OxREP, that approach was bound to fail for the kinds of reasons discussed by Ahluwalia and Patel. But a process based on the ideas of concerted unilateralism seems to be offering a way forward. Indeed, what is happening seems to be a perfect example of the process of concerted unilateralism which we have been describing in this paper. As in other policy areas, this process means first reaching agreement on the rules of the game—here an obligation on each country to make a ‘fair’ contribution to the global policy objective to limit global warming to +1.5°C—and then allowing a (gradual) process to work itself out in which each participating country works out how and when to make its own unilateral contribution.

The distributional nature of the global climate-change problem introduces important difficulties beyond those which we have just discussed.

Current WTO rules will likely require reform to accommodate measures that support decarbonization and a just energy transition. Those that discriminate against countries which are not operating appropriate policies aimed at decarbonization—like border carbon adjustment (BCA) taxes—will become important. These will be difficult to implement in a non-protectionist manner. A new typography that differentiates between ‘good’ subsidies to reduce emissions and those that are solely to support domestic industry is already under discussion at the WTO. Here, emerging-market economies will face difficulties, given they will be slower to either develop or adopt low emissions technology (e.g. they will not be able to afford to adequately invest, at least in the required time frame). There will thus need to be some kind of global fund to compensate those countries for the effects of these policies and—at the same time—to subsidize action by countries that do pursue adequate decarbonization policies.

Where will this fund come from? The World Bank and other international financial institutions will need to find ways to make the necessary loans. And the international monetary system will need to find ways of supporting the balance of payments deficits which these emerging market economies will need to run in order to undertake the high levels of private-sector investment which will be necessary as these new technologies are adopted (see McKibbin and Vines, 2023).

Issues for the World Bank: supporting lending in support of global health policies

Health externalities, and the need to insure against future pandemics, is another issue which has significant implications for the World Bank. Until now, the institution has been focused on a combination of knowledge and finance for the infrastructure development in the world’s poorest countries. It now needs to focus on lending to help deal with the major global externalities to do with the risk of, and response to, global pandemics (Brown and Susskind, 2020; Susskind and Vines, 2020).

This will require a significant change of practice for an institution which has until now been focused on country-by-country need for infrastructure and for institutional development (Morris, 2023). Lending by the World Bank and other international financial institutions will be necessary, given what we know about how hard it is to attract private funding in sufficient quantity to support the provision of public goods. There will need to be significant subsidized rates to many of the world’s poorest economies, and emerging-market economies.

VII. The rise of a new hegemon

Perhaps the most important shock to the global economic order is that it has once again undergone an enormous jolt to its leadership, this time due to the rise of China (see Mitter, 2024; Evans, 2024; Garnaut, 2024). Until recently, the world was unipolar and the US was able to lead. Now, global power is in the hands of (at least) two dominant countries. Furthermore, the interests of other countries, particularly those in the Global South, matter more than ever (see Stuenkel, 2024).

(i) Complexity and indeterminacy

The simple addition of an emerging rival to the existing hegemon, in the form of a new leader, introduces several destabilizing new strategic interactions: between the two leaders (see, for instance, Helen Milner and Erik Voeten (2024, this issue) on the US and China), between each leader and their followers (see, for instance, Fried (2024) on the US and Canada), between each leader and the other leader’s followers (see, for instance, Evans (2024) on China and Australia; Jayakumar (2024) on China and Singapore), as well as between the followers themselves (see, for instance, David Miliband (2024, this issue) on Britain and the rest of the world; Stuenkel (2024) on the Global South and the rest of the world). These interactions are depicted in a schematic manner in Figure 3 below.

In addition to the complexity of existing international arrangements, it is also important to note its indeterminacies—what Milner and Voeten (2024) call ‘international regime uncertainty’, and what Rodrik and Walt (2024) describe as a state of ‘flux’ and ‘transition’. For both these papers, this indeterminacy is the key feature of the global economic order that must be understood: the former takes it for given, exploring how it influences international ‘economic and political activities’; the latter asks how it can be resolved, beginning with a conceptual blank sheet of paper and proposing a ‘meta-regime’ of minimal agreement among countries as a first step towards a new global economic order. We return to both these arguments in more detail in a moment.

But what exactly are the indeterminacies in the prevailing pre-system? To begin with, there is an unresolved tension between one leader, the established US hegemon, and another rival leader, the rising Chinese hegemon. In this case each leader also has its own changing ‘club’ of followers which subscribe, only in part, to the rules of the game chosen for that club. In turn, followers also now move far more fluidly between clubs, picking and choosing which particular rules to follow (what Miliband (2024) calls a ‘multi-aligned’ world, where countries ‘vary in their alignment from issue to issue’). This final feature is crucial; there is an indeterminacy in the global economic order at present, and a sense that there are rival possibilities competing with one another. In the Young set-up, it is as if the old rules of the game—the ‘non-system’—have now broken up, and we have found ourselves back in the ‘meta-game’ of the two-stage game, and a sort of ‘pre-system’ now prevails once again. The consequence, though, is that the world, as Miliband (2024) puts it, is ‘not just disordered; it is becoming more dangerous’.

Furthermore, it appears that the world is dividing into not just two but many clubs or blocs: not just a US bloc and a China bloc but also a European bloc sitting on the edge of this divide. Indeed, the European Union has created a model of how particular nation states might work together as a bloc—achieving a result in which the outcome is greater than the sum of the parts— and thereby doing something which enhances Europe’s potential for global leadership (see Clunies-Ross, 2005). Furthermore, there is also now a large ‘uncommitted’ bloc as well, which we can call the Global South (see Stuenkel, 2024). In such a world, in which economies of scale and network externalities are important, both the US and China are likely to be tempted to support countries which join their own bloc and penalize those that do not.⁴² This is an economic interpretation of the motivation behind the ‘friend-shoring’ policies upon which the US has embarked.

There is much at stake. In an important new paper, Laurence Kotlikoff and a group of colleagues present a set of model simulations which predict a future for the world economy that is very different from the world that we can see around us at present. The baseline projection of their model suggests that by 2100 China and India will become the world’s largest two economies, with, respectively, 27.0 per cent and 16.2 per cent of world GDP, respectively. At this time, the shares in world output of the US and Western Europe (including the UK) are projected to be 12.3 per cent and 11.9 per cent. By contrast, in 2017 (which is the starting date for their modelling exercise), the US and China each accounted for roughly 16 per cent of world GDP, lagging behind Western Europe (plus the UK) which accounted for almost 20 per cent. And India produced only 7.0 per cent of world output. So there appears to be a lot to play for (see Kotlikoff et al., 2023).

(ii) Security

The security dimension greatly complicates the strategic interactions between the US, China, and other nations depicted in Figure 2. The kind of friend-shoring described above will inevitably spill over into defence treaties and defence partnerships. The global order will need to develop a system of international relations management which addresses these defence-related friend-shoring tensions. This question is discussed in some detail in Jayakumar (2024).

More generally, defence and security expenditures, while foundational for the state and politics, have negative effects on productivity growth. When unemployment is the dominant concern, expenditure on defence raises demand for labour like any other public expenditure. At full employment, it becomes a drag on growth in productivity and incomes in a country that lacks comparative advantage in the production of defence materials. Demands for higher current defence expenditures must be justified by the value they contribute to defence, and not by some falsely asserted contribution to productivity growth.

Potentially even more costly for long-term economic growth are calls for restrictions on trade with strategic competitors—obviously China in the current discussion—on grounds that open trade will expand the competitor’s economy and future defence capacity. For example, the US, EU members, and others are actively discussing further restricting trade and investment with China and some other countries in goods and services that are important for national security. Well and good. But the costs are high without corresponding defence benefit when trade restriction is extended beyond the ‘small yard, high fence’ surrounding commodities the open trading of which really does damage national security interests (Sullivan, 2023). The boundaries of the fences need to be assessed rigorously, with realistic calculation of genuine defence benefits and economic costs (see, for instance, Evans (2024) for an example of how to engage in this difficult balancing act in the context of Australia and China).

Then there is the dual-use argument. Advanced technologies are now necessary for high-end defence reasons, and it is important to prevent China from obtaining these. Security risks are posed to the US, it is said, by Chinese behaviour in the South China Sea, and China’s ambitions in relation to Taiwan. Indeed, these are said to present dangers not just to the US, but to the stability of the global economic and political system. It is important, it is said, for US policy to impede this behaviour and these ambitions, by ‘friend-shoring’ (sourcing production inputs from suppliers in allied countries) in order to secure access to critical production inputs. The US Chips Act is an example of such a policy. Yellen (2022) presents a carefully calibrated version of this argument. Allies of the US are also already facing these choices. There is broad agreement in the West on the need for reliable supply—e.g. the Critical Minerals Partnership agreement—so this is not just an issue for the US.

(iii) Values

There is, furthermore, an additional argument that is the most broad-reaching of all and complicates the strategic interactions between the US and China in a fundamental way. This is the fear that the hegemonic rise of China will put the liberal international order at risk because of the failure of China to respect liberal values.

One can see evidence of this argument in the fear manifested by many of our US colleagues about their colleagues in China, who are, for example, now no longer able to participate in online international discussions for fear that anything which they say of a frank and critical nature will be used against them by the political authorities. The risks of a world order in which a Chinese hegemon has leaders who do not share liberal values, and exercise their power autocratically, and without fear of challenge, is now very great. We have all seen what has happened in Hong Kong. One is now under an obligation, it is said, to use economic policy to impede the progress of China, to diminish the risk of an outcome in which an authoritarian China is the global hegemon. That obligation makes clear why—to those who are proposing them—the anti-Chinese aspect of the policies in the IRA should not be resisted but instead should be actively supported.

(iv) Strategy

Many observers argue that the move towards economic nationalism and the rise of China—which have gone together—will not lead to a good outcome. What happened when Germany challenged Britain in the run-up to the First World War is taken to be a lesson about what might emerge from the challenge which China is mounting to the US. See Mearsheimer (2018) and Mearsheimer and Rosato (2023). It is our hope that some kind of multilateral global order can be established to guard against this kind of danger, and it is our belief that such an outcome will in fact be possible. A number of the papers in this issue of OxREP bear directly on whether we are right to hold this view.

Certainly, China’s power and ambition threaten the post-Cold War vision of an essentially unipolar, US-led liberal internationalist order which we depicted in Figure 1. But what will take its place is not, in our view, an essentially unipolar, Chinese-led authoritarian order which will threaten liberal democracy in countries like our own. It is unlikely that China alone, or with its authoritarian allies, will come to dominate the world, because there are too many other strong powers. The most likely outcome is a multipolar order of great powers, a simplest version of which—with only two great powers—is depicted in Figure 2. Each of these powers will dominate their ‘near abroad’—as the US dominates the Americas—and each of them will take a broadly equal role in creating a collective global leadership. This is, in fact, pretty much what the architects of the post-Second World War order in 1945 had in mind; it was reflected in the structure of the United Nations Security Council. In this new order, East Asia and the Western Pacific will lie within the Chinese sphere of influence.

Of course, such an outcome is not inevitable. We are fortunate to have a short paper by the former Australian Foreign Minister, Gareth Evans, building on a speech he gave in September 2023, in which he discusses what is necessary in order to reach such an outcome. Evans (2024) begins by emphasizing how much will be put at risk, and how devastating the negative impact will be, if the geopolitical environment blows up: if the strategic competition between the United States and China—that is, the struggle for primacy—becomes even more fierce. The ‘overwhelming need right now,’ he says, ‘is the de-risking of the whole regional geopolitical environment.’

Dani Rodrik and Stephen Walt agree with this. At the beginning of their paper (Rodrik and Walt, 2024) they write as follows:

If any of the major powers make economic and geopolitical predominance their overriding goal and lesser states make relative power their chief concern, the prospects for a more benign global order are slim. But major powers such as the US and China can remain secure from external threats while enhancing the overall well-being of their societies at the same time—by preserving the physical conditions necessary for human existence, advancing economic prosperity, and minimizing the risks of major war.

If, as they believe, global primacy is an unreachable goal, then its all-out pursuit is counter-productive. It will buy little in terms of national security. But, at the same time, the pursuit of this goal will make it harder to achieve other important economic, social, and environmental goals that can only be achieved by means of international cooperation. As a result, Rodrik and Walt argue, major powers, including the US and China, will need to act in a different way, and make different kinds of choices.

Evans gives a short clear description of what these choices might need to be, for each of the US and China. For the United States, says Evans, the restraint that is needed above all is to step back from demanding recognition of its continued primacy. Washington should do more ‘to accommodate the reality that China . . . is no longer prepared to be . . . just a rule-taker, but is determined to be a major player, a participating rule maker’. His call for the US to step back ‘both regionally and globally’ is consistent with the story which Rodrik and Walt, and we ourselves, are telling. The US must choose, says Evans, not to view China’s growing strength as a threat to US unipolar hegemony. It should, instead, see China’s ambitions as geographically limited and should be prepared to make the concessions which are necessary to enable China’s objectives to be fulfilled, and to do this within a world which is multipolar rather than unipolar.

The multipolar order which all of us think of as desirable would not be as congenial for us in the West as the unipolar order has been. But it should be possible for us in the West to protect our political values, and preserve the political systems that support them, in this multipolar world—just as authoritarian countries will be able to preserve their values and systems. Hence, we disagree with the argument that it is necessary to preserve some kind of a unipolar order in order to defend liberal democracy in Western countries. Nevertheless, we think that it is extremely uncertain as to whether the US will act in an appropriate manner, given the domestic political turbulence within that country.

And what might be required of China? ‘From China’, says Evans, ‘some restraint is also unquestionably required . . . Its recently reasserted territorial claims over the South China Sea . . . are indefensible in international law, as has been its militarization of certain contested reefs and islets.’ Evans notes that China has clearly broken its 1997 treaty pledge to the UK to respect, at least for 50 years, Hong Kong’s governance system. Other unhelpful behaviour has been the way in which, in international fora, China has recently played a much more active spoiling role on human rights than it did in the past. We are fortunate to have in this issue a paper by Shashi Jayakumar (2024) which carefully documents the kinds of threat which such spoiling activity now presents to the prospects of global order.

Mitter (2024) provides an important insight as to whether China might choose to participate in the kind of outward-looking multilateral order which we hope to see sustained. Mitter describes the astonishing reduction in the working-age population in China which is likely to occur by 2100. He talks about how immigration might possibly fill some of that gap, but concludes that any response of this kind is unlikely to be quantitatively important.

Much more significant has been—and will be—the ambition of the Belt and Road project, and its successor, the Global Development Initiative, to connect Chinese business with the labour force in other countries. Chinese authorities will have to deal with the fact that, with a falling population, there will be diminishing opportunities for profitable domestic investment within China, simply because of the falling rate of expansion of domestic markets, due, in turn, to the fall in the level of the population. In the face of this likely outcome, it seems that those saving in China, and the Chinese financial system, will need to go on investing abroad—even more, perhaps, than in the recent past. That high level of foreign direct investment (FDI) will need to lead, initially, to a high level of Chinese exports so as to sustain high levels of aggregate demand within China, in order to make up for the lack of investment in capital equipment at home. But more than this, that FDI will need to create Chinese-owned assets abroad which, in due course, generate an income returning to China, so as to make it possible to meet the consumption needs of an increasingly elderly population. Very simply, one can think of the Belt and Road, and the Global Development Initiative, as a set of policies designed to connect Chinese investors with a larger labour force.

To us, it appears that these macroeconomic implications of a falling workforce in China might well propel Chinese policy-makers towards a support for a workable multipolar global order, simply in order to enable Chinese economic growth to continue.⁴³ Nevertheless such an outcome is by no means guaranteed. Mitter (2024) makes this very clear, in what is a careful and systematic account of the policy choices that the Chinese authorities now face.

A world in which the US and China make the kinds of choices identified by Evans—and by Rodrik and Walt and Mitter—would differ significantly from a world in which these two major powers constantly strive to maximize their relative power. The paper by Rodrik and Walt begins to describe the ways that things might change. In particular, they describe a kind of meta-regime for cooperation which might be built in such a world: a framework in which it would be possible to structure conversations about matters of shared concern. Even where agreement about a particular set of policies initially proves impossible, the objective of such a set-up would be to enhance communication among the relevant parties, to clarify the reasons for disagreements, and to create circumstances which might facilitate either eventual agreement, or accommodation, if that is all that can be achieved. The creation of an environment such as this might incentivize the US and China—and other states as well—not to inflict unnecessary harm on each other as they act independently in their own interest. And, in favourable circumstances, such an arrangement might gradually lead to significant increases in cooperation—even among adversaries—as participants in the meta-regime build trust between themselves.

Rodrik and Walt go on to illustrate the practical implications of such a regime by applying their ideas to several different policy domains—some of which are analysed by other authors in this OxREP issue. And they specifically examine how the existence of such a regime might help to moderate the competition between and the US and China.

VIII. Conclusions

It is now time to draw our paper to a close. We go back to where we began. How can we summarize what we have learned from the engagement between economists and international relations scholars that is displayed by the papers in this issue of OxREP?

We have been fortunate, when assembling this issue, to be able to bring together a large number of extraordinarily interesting papers, written by a very talented group of people, spanning a wide range of topics in international relations and international economics. These papers provide an overall account of the challenges which the world now faces and the steps which must be taken to meet these challenges. There is a sense in which such an overall account encompasses the particular pieces of analysis in each of the papers. For this issue of OxREP, the whole is much bigger than the sum of the parts.

(i) Framing

We described in the introduction how the papers by the IR scholars in this issue can be used to frame our understanding of the processes of cooperation in policy-making about particular economic policies.

As already noted, the paper by Hurrell (2024) has made it clear that the complexity of the world economy is now forcing us to think again about the foundations of international cooperation in economic policy-making. Hurrell wishes to argue for the importance of ‘deep pluralism’ as we perform this re-examination. Partly this need comes from the heterogeneity of interests and beliefs that is now evident across many countries. But, more than this, it comes from the way in which increasing interaction over time, between more and more countries, is producing what are, effectively, different ways of looking at the world. These include different ways of looking at the problems which we all face together, such as global warming and global pandemics.

Clearly, the rise of other significant countries, especially in the Global South, has made this re-examination much more challenging. The task of cooperation is more difficult than, say, it was at the Bretton Woods Conference, because of differing values and interests among many different nations; the essay by Stuenkel (2024) makes this clear. And the increase in global uncertainty, which is discussed in the paper by Milner and Voeten (2024), has also made cooperation much harder.

(ii) Broad policy regimes and particular policy proposals

Very broadly, as we have argued in this paper, IR scholars are inclined to analyse international economic-policy regimes, whereas economists tend to study particular policy proposals. Of course, this is a generalization. But it is striking how robust this distinction turns out to be—in the formal papers written for this issue, the informal conversations that have supported it, and the wider literatures that our contributors have drawn upon.

The cooperation concerning particular policies is described in great detail—policy by policy—in the papers by the international economists in this OxREP issue. In particular, the paper by Obstfeld provides a detailed account of the global macroeconomic policy set-up, beginning with the Bretton Woods regime and coming forward to a discussion of what needs to be done in macroeconomic policy now, and going into the future. The paper by Guzman and Stiglitz does the same for the policy framework which has governed the international trading regime, and discusses in detail the kinds of choices which now need to be made in that policy-making area. The papers by Garnaut and Ahluwalia and Patel describe, in detail, policies which are being adopted in relation to climate-change issues in two key countries, namely China and India. By contrast, the papers by IR scholars in this issue describe how these macroeconomic and trade-related policies came to be fitted together in an overall policy regime in the immediate post-war period, and go on to examine how this ‘fitting together’ continues to be necessary at a time when climate-related policies, and health-related policies have been added to the mix of policies that are necessary.

This difference in perspective is most evident if we consider the two papers by John Ikenberry (2024) and Jonathan Kirshner (2024). The former provides a persuasive account of what he calls the ‘logic and character’ of the liberal international order which emerged after the Second World War: an ‘interconnected “system” of layered rules, institutions, bargains, and projects’—a sort of geopolitical environment or ecosystem—in which states could survive, protect their values, institutions, and interests, and manage the complex problems of economic, political, and security interdependence. Ikenberry goes on to write as follows:

It is possible to identity five types of ‘problems’ that liberal democracies have grappled with over the last centuries, particularly in the twentieth century, during the period of American dominance. These are the problems of anarchy, hierarchy, interdependence, liberal openness, and geopolitical vulnerability. (Ikenberry, 2024)

And the body of Ikenberry’s paper is devoted to arguing that the post-war liberal international order was successful because liberal democracies, led by the US, were particularly capable when it came to dealing with these five ‘problems’. Kirshner then argues persuasively that this capability depended on the kind of embedded liberalism to be found in post-war economies, which we described above.

Our view, as economists, of how the post-war liberal international order emerged is complementary with their view. We see it as having emerged as a result of four features of the post-war global economic policy-making regime which we have discussed—the PIPM. The global order of liberal internationalism, described by authors like Ikenberry and Kirshner, required some form of successful economic management. The kind of international economic analysis which we have carried out in our paper describes the sorts of economic policies which we think were necessary in order for a well-functioning global economic order to come into existence and to be sustained. Such a policy regime was pioneered at the beginning of the post-war period, and it was transformed, but preserved, with the emergence of the global non-system, in the ways that we have described. Nevertheless, that policy regime is now in the process of disintegrating, for the reasons which we have been discussing. We believe that this collapse of the existing system of global economic management is what is in the process of causing the disintegration of the post-war liberal international order. This is why the Great Moderation has been turning into the Great Unravelling.

(iii) Concerted unilateralism: a new analytical framework

Standing back from this point of view about current circumstances, we have been led, in our own paper, to provide a new analytical framework for thinking about the intentional cooperation in economic policy-making which we have been discussing. We have been prompted to do this precisely because of the collaboration between IR scholars and international economists which we have described. Of course, any new framework will build on the existing work of IR scholars and international economists. But that work has been dominated by the idea that international cooperation in economic policy-making involves the solution to a prisoner’s dilemma problem, a solution which is forever in danger of unravelling, due to the existence of free-riders. This idea turns out to be profoundly unhelpful when it is confronted with the historical narrative we have presented in our paper. We have been able to go beyond that conceptual difficulty—going at least part of the way that is necessary—by deploying Peyton Young’s path-breaking analysis of the evolution of conventions and norms in support of economic cooperation (Young, 1998b, 2024).

We have used Young’s ideas to illuminate, in a new way, the history of cooperation in economic policy-making since the Second World War, its present state, and its likely future development. In particular, we argue that international cooperation about economic policy-making involves a two-stage process, to which we give the label ‘concerted unilateralism’. In the first stage of this process the rules of the game are chosen and the relevant GEI(s) formed; this is the ‘concerted’ stage. Then, in the second stage, the game is played, and each country seeks its own advantage; this is the ‘unilateral’ stage. In this second stage each team does not cooperate with the other team(s) but, instead, sets out to pursue its own advantage. Nevertheless it does this following the rules of the game that have been established in the first stage. We have developed this analytical framework in a way which has not been done before. Our idea of concerted unilateralism provides a much-needed change to the conventional way of thinking about cooperation in international economic policy-making: it puts to one side the idea that such cooperation involves the solution of a prisoner’s-dilemma problem.

Viewed this way, it is clear why the complex task of international cooperation has become increasingly difficult. There are many players, each with different values, all of whom have competing interests, and all of whom are now coming to challenge the hegemonic position of the United States. And there is a new competing hegemon—China. And others, like India, are soon to follow China in this way. We have argued that, because of these complications, we are now back in a meta-game, where the rules of the game, going forward, have not yet been determined, and the contract, or contracts, that govern the global economic order have not been agreed.

(iv) Looking ahead

Nevertheless all is not lost. The over-arching challenge is to conserve the proven wisdom from the past, while taking account of new circumstances requiring policy adjustment. Economic policy has been thrust into a brave new world before the ideas on how it works, and on how policy works within it, have been properly settled. The current disorder presents us with a new challenge: not the known risks of the non-system that followed the collapse of the Bretton Woods system, but the unfamiliar uncertainties of the pre-system we find ourselves in today. This is a rise of ‘Knightian uncertainty’ in the global economic order, to which Milner and Voeten (2024) rightly call attention. At this time, we will need innovative analysis, of the kind which economists have been capable in the past, to build on the wisdom of experience.

Furthermore as Jonathan Fried points out in section III of his paper, there is much granular plumbing in place to do with international governance, which will be hard to dislodge. The fact is, he says, that a good deal of international cooperation is alive and well, behind and below the headlines. It is worth reminding ourselves of the international institutions which Fried goes on to list as part of this plumbing: the International Maritime Organization, the International Civil Aviation Organization, the Universal Postal Union, the International Telecommunications Union, the World Meteorological Organization, and the Bank for International Settlements. These institutions fulfil a necessary coordinating role across a very wide range of international activities and linkages. There are an extraordinarily large number of other such institutions whose activities are unlikely to be swept away by the geopolitical concerns which we have been describing.

Politics will be important, as well as economic policies. For Jonathan Kirshner, a nation—and in particular the US—is only able to participate helpfully in global governance—and, in the case of the US, is only able to be a reliable hegemonic leader in the process of global governance—if some form of embedded liberalism remains in place. For the US this means that much rebuilding will be necessary. It is Kirshner’s considered view that this rebuilding will not be possible. It is John Ikenberry’s hope that it might be. And it is our hope that good analysis of economic policies, and good economic policies themselves, might help to tip the political balance in the right direction.

There is much to be understood, and then there is much to be done. It is the prospect that we all might be able to do this together—both economists and political scientists—which gives us hope that Ikenberry might be right.

We are grateful to the many contributors to this issue of the Oxford Review of Economic Policy (OxREP) for the discussions we have had with them about the topics analysed in this paper, and we are especially grateful to Montek Ahluwalia, Jonathan Fried, Ross Garnaut, and Maury Obstfeld for the detailed comments which they provided on an earlier draft of the paper. The commentators on other papers in this OxREP issue, and the referees for those papers, have also provided much useful insight. Our fellow editors of OxREP made valuable comments on an early draft of the paper, particularly Christopher Adam, who gave us guidance on numerous occasions, and Emily Jones, who challenged us to say more about what we have learned from IR scholars. We would like to thank Shiro Armstrong, Olivier Blanchard, Peter Drysdale, David Enoch, Martin Parkinson, Heather Smith, John Tasioulas, Richard Vines, and Hugh White for many suggestions, and we have also received helpful comments from an anonymous referee. The insights which we have gained from Peyton Young will be obvious to anyone who reads the paper.

References

Footnotes