Bretton Woods for a Geoeconomic Age
By Matthew Parish
Wednesday 23 September 2026
The conference at Bretton Woods in July 1944 took place while the Second World War was still being fought. The Allied victory in Europe remained almost a year away, the war against Japan would continue longer still and much of the world’s industrial infrastructure had been destroyed. Nevertheless, representatives of forty-four countries assembled in a hotel in New Hampshire to discuss something that might have appeared premature: the architecture of the international economy after the war.
Their foresight was extraordinary. The architects of Bretton Woods understood that the economic catastrophes of the interwar years had contributed to the political catastrophes that followed them. Competitive devaluations, protectionism, sovereign financial crises, unemployment and the disintegration of international commerce had not merely impoverished countries. They had helped poison relations between them, until economic disorder became political disorder and political disorder culminated in war.
The institutions that emerged from Bretton Woods were therefore based upon a profound insight. International economic liberalism could not survive merely because economists believed it efficient. It required institutions capable of reconciling international openness with the legitimate political requirements of sovereign states. That problem has returned, and although the circumstances are different the intellectual challenge is remarkably similar. The twenty-first century may need another Bretton Woods.
The end of innocent globalisation
The economic order that followed the Cold War rested upon an optimistic assumption. As countries became increasingly integrated through trade, investment, finance and technology, geopolitical rivalry would become progressively less important. Economic interdependence would moderate political conflict because countries would acquire too much to lose from confrontation.
There was considerable truth in this proposition. Globalisation generated enormous wealth, brought hundreds of millions of people out of extreme poverty, distributed sophisticated technologies across continents and created supply chains of extraordinary complexity. Companies optimised production across borders, financial markets allocated capital internationally and consumers acquired access to goods manufactured almost anywhere on Earth. For a time, it seemed that economics might gradually subordinate geopolitics.
Yet the system contained an assumption that became increasingly difficult to sustain: that economic dependence could be considered separately from political power. Russia’s full-scale invasion of Ukraine demonstrated the problem dramatically. European countries discovered that inexpensive Russian hydrocarbons had created strategic vulnerabilities, while Russia discovered that her integration into Western financial systems exposed her to sanctions. China observed both developments closely. The United States increasingly regarded Chinese dominance of certain manufacturing supply chains as a national-security problem, while China increasingly regarded her dependence upon American semiconductor technologies and dollar-based financial infrastructure in precisely the same way.
The problem is therefore no longer whether geoeconomics should exist, because it already does. Governments have rediscovered that control over energy, finance, technology, shipping, minerals and manufacturing can become an instrument of political power. The question is how to prevent this rediscovery from destroying the international economic system that generated so much prosperity in the first place.
Hayek’s warning and Keynes’s answer
The first two essays in this series considered what Friedrich von Hayek and John Maynard Keynes might have thought about contemporary geoeconomics. Their respective insights provide useful intellectual foundations for a new institutional settlement because each identifies a danger that the other tends to underestimate.
Hayek would warn against giving governments broad powers to redesign international production. No ministry possesses enough information to determine the optimal geographical distribution of semiconductor factories, battery plants, artificial-intelligence infrastructure, steel mills, pharmaceutical production and energy supplies. Once governments begin subsidising strategically favoured industries, lobbying may replace competition and political relationships may begin to determine the allocation of capital.
Keynes would reply that markets cannot be expected to price national survival correctly. A country that depends entirely upon a hostile power for ammunition, energy or critical technology may discover that the cheapest supply chain is also the most dangerous. Some redundancy, domestic productive capacity and diversification may therefore be worth purchasing even at considerable cost.
Both arguments contain essential truths, and the purpose of a twenty-first-century Bretton Woods should be to construct institutions around their uncomfortable coexistence. Markets should remain presumptively free, but governments should possess recognised mechanisms for addressing genuine strategic vulnerabilities. The central institutional problem is distinguishing those vulnerabilities from ordinary protectionism.
That distinction cannot be left entirely to individual governments because every government has an incentive to describe politically influential domestic industries as strategically essential. If steel is strategic, automobiles may become strategic because they require steel. Batteries then become strategic because electric automobiles require batteries, lithium becomes strategic because batteries require lithium and ports become strategic because lithium must be imported. Banks can then become strategic because ports and mines require finance. Eventually everything is national security and nothing is free trade.
A revised international economic system therefore requires rules governing geoeconomics itself. Those rules must recognise genuine security interests without permitting national security to become the universal incantation by which governments escape the disciplines of international commerce.
A Geoeconomic Security Council
One new institution might be called a Geoeconomic Security Council. Its purpose would not be to determine national-security policy, because sovereign countries would retain the right to decide what they consider necessary for their defence. Instead, the Council would provide a forum in which measures adopted in the name of economic security could be notified, explained, challenged and periodically reviewed.
Governments imposing export controls, strategic tariffs, investment restrictions or major industrial subsidies would explain the security rationale behind them. Other countries could contest measures they regarded as disguised protectionism, while independent economic analysis could estimate their consequences for international markets and supply chains. The exercise would not eliminate disagreement, but it would force governments to identify the relationship between a claimed security risk and the economic intervention ostensibly designed to address it.
The institution should not possess an absolute veto over national-security measures. No Ukrainian government should require permission from an international organisation before restricting Russian ownership of Ukrainian defence infrastructure, just as no country can reasonably be expected to surrender ultimate control over matters essential to her survival. Nevertheless, transparency matters because exceptional measures have an unfortunate tendency to survive the emergencies that justified them. Governments should periodically have to explain not only why an intervention was necessary when introduced but why it remains necessary years later.
The objective would therefore be to establish an international presumption that geoeconomic interventions are exceptional rather than normal. National security would remain a legitimate justification for intervention, but it would not become a phrase that ends the discussion as soon as it is uttered.
Mapping strategic dependencies
A second component would address strategic dependence directly. The traditional international economic system concentrates upon trade flows, tariffs, currencies and financial balances. A new system would also need to measure concentrations of dependency, because the geographical concentration of an essential supply can matter as much as its price.
Suppose 80 per cent of an indispensable pharmaceutical precursor comes from one country, virtually all advanced semiconductor fabrication occurs within a small geographical area or a critical mineral is overwhelmingly processed by one state. Alternatively, suppose international cloud infrastructure depends upon a handful of companies operating under the jurisdiction of a single country. These are economically different situations, but they share a common characteristic: concentrated dependency creates the possibility of disruption or coercion.
A new international institution could maintain continuously updated maps of critical supply chains. Governments would identify categories of goods and infrastructure whose sudden disappearance would cause severe economic or security disruption, while independent analysis would assess concentration risks. Private companies would necessarily contribute information, although commercial confidentiality would have to be protected.
The purpose would not be to eliminate dependence, because that would be both impossible and undesirable. International trade exists precisely because countries specialise. The objective would instead be to distinguish ordinary interdependence from dangerous concentration. A country purchasing lithium from ten politically diverse suppliers has an international dependency but not necessarily a strategic vulnerability. A country obtaining an indispensable defence component from a single potentially hostile supplier has both, and that distinction ought to sit at the centre of twenty-first-century trade policy.
Diversification before reshoring
Such an institution would encourage another important principle: diversification should normally be preferred to autarky. Contemporary political rhetoric frequently assumes that economic security means bringing production home, yet reshoring is only one way to reduce vulnerability and frequently the most expensive one.
Consider a European country dependent upon a single foreign supplier for a critical mineral. She could attempt to mine the mineral domestically, assuming appropriate deposits existed, or she could develop supply relationships with Canada, Australia, Brazil, African producers and other politically independent sources. The second approach may provide greater resilience at substantially lower cost, while preserving the advantages of international specialisation.
This is one area in which the international system can reconcile Hayek with Keynes. Keynesian concerns about security justify reducing dangerous concentrations, while Hayekian concerns about planning suggest allowing markets to discover alternative suppliers wherever possible. The objective should therefore not be national self-sufficiency but international redundancy: a system in which countries remain dependent upon one another without becoming dangerously dependent upon any single counterpart.
A resilience facility
Poorer countries create a further difficulty because diversifying supply chains, constructing strategic infrastructure and maintaining emergency reserves cost money. Wealthy countries can subsidise semiconductor factories and accumulate energy reserves, whereas poorer countries frequently cannot. If strategic resilience becomes something only wealthy states can afford, the new geoeconomic order will deepen rather than reduce international inequality.
A revised World Bank might therefore acquire a substantial new resilience mandate. The institution was originally designed to finance reconstruction and development, but in the twenty-first century development increasingly requires resilience against geopolitical and supply-chain shocks. Financing could support ports, electricity interconnections, digital infrastructure, alternative transport corridors, mineral-processing facilities and diversified energy systems.
This would also address one of the most dangerous features of contemporary geoeconomics: the possibility that poorer countries become merely objects of competition between wealthier powers. A country possessing strategically important cobalt, lithium or rare earth deposits should not be forced to choose between becoming economically dependent upon China and becoming economically dependent upon the United States. International development finance could help her construct infrastructure, processing capacity and diversified commercial relationships so that she retained greater political independence.
The geopolitical objective of development policy should therefore not be to purchase allegiance but to make allegiance less economically compulsory. A genuinely liberal international order should enlarge the choices available to weaker countries rather than compel them to enter the economic sphere of whichever great power offers the largest subsidy.
The IMF as insurer against economic coercion
The International Monetary Fund would also require a revised mandate. The traditional IMF intervenes principally when countries experience balance-of-payments and financial crises, yet future crises may increasingly have geopolitical causes rather than arising primarily from domestic macroeconomic failures.
A country may suddenly lose access to an export market because another government wishes to coerce her. Her banks may be excluded from a financial network, her energy supplies may be interrupted or foreign investors may withdraw because geopolitical tensions have abruptly changed their assessment of risk. These events can create financial crises even where the victim country’s macroeconomic policies were reasonably prudent.
A revised IMF could therefore establish an Economic Coercion Facility through which countries suffering severe externally imposed economic disruption could receive temporary liquidity while they reorganised trade and finance. Assistance plainly could not compensate every government affected by sanctions, and eligibility would require careful institutional assessment of the circumstances, including the legality and international context of the measures concerned.
The underlying principle would nevertheless be important because economic coercion derives much of its effectiveness from the fact that adjustment takes time. A country deprived suddenly of a major export market may eventually find another, just as a country deprived of an energy supplier may eventually build alternative infrastructure. International financial insurance could provide the time necessary for that adjustment and thereby reduce the political leverage created by economic dependence.
The WTO must learn about security
The World Trade Organization presents perhaps the most difficult institutional challenge. The post-war trading system was constructed around progressive reduction of discrimination and barriers, whereas contemporary geoeconomics moves in the opposite direction. Governments increasingly distinguish between friends, adversaries and politically ambiguous trading partners.
Pretending that this is not happening will not preserve the trading system; it will merely make the WTO progressively less relevant. A reformed WTO therefore needs a more sophisticated security architecture in which governments retain room to restrict trade genuinely affecting national security, but security exceptions do not become unlimited licences for protectionism.
One possible organising principle would be proportionality. A government invoking national security would not merely ask whether an industry had some conceivable security significance. It would also ask whether the particular restriction was reasonably related to the identified vulnerability and whether a less trade-restrictive measure could address substantially the same problem.
A strategic stockpile might sometimes be preferable to a permanent tariff, while diversification might be preferable to an import ban. Inspection requirements might be preferable to excluding an entire country’s products, and restrictions upon particular military technologies might be preferable to broad technological separation. Security would remain legitimate, but security policy would acquire disciplines designed to prevent exceptional restrictions expanding indefinitely into ordinary economic nationalism.
Sanctions and their contradictions
Sanctions present another issue that the original Bretton Woods architects could scarcely have anticipated in their contemporary form. Financial sanctions have become one of the principal instruments of modern statecraft because global finance is highly interconnected, and their effectiveness derives precisely from the success of international economic integration.
Yet sanctions contain a structural paradox. The more frequently financial networks are used as instruments of geopolitical coercion, the stronger the incentive for targeted countries to construct alternatives. Russia has experienced extensive exclusion from Western financial mechanisms, China has obvious reasons to prepare herself for the possibility of similar measures in a future crisis and other countries observe these precedents when deciding how much of their financial infrastructure should depend upon systems ultimately subject to decisions taken in Washington or Brussels.
This does not mean that sanctions are necessarily wrong. In some circumstances they may represent a preferable alternative to military force, particularly where they respond to grave violations of international law. It means that sanctions have systemic costs beyond the immediate economic damage imposed upon their targets.
A revised Bretton Woods system should therefore encourage clearer principles governing financial sanctions, particularly secondary sanctions affecting third countries. Measures responding to aggression, proliferation or other grave breaches of international obligations stand on a different footing from routine attempts to use financial infrastructure to extract commercial or diplomatic concessions. If every major power weaponises every network she controls, eventually there will be fewer genuinely common networks left.
A neutral payments layer
This leads to perhaps the most technically important element of a new Bretton Woods settlement: an internationally interoperable payments architecture. Cross-border payments remain considerably more expensive and cumbersome than domestic payments, while geopolitical fragmentation increasingly threatens to divide payment infrastructure itself.
The solution need not be a single world currency, nor does it require abolishing the dollar, euro, renminbi or other national and regional currencies. Instead, central banks could develop common technical standards allowing different payment systems to communicate with one another. Digital currencies, conventional bank money and future tokenised financial instruments could operate through interoperable settlement protocols without requiring political agreement upon a common currency.
The distinction is important because a common currency requires deep political integration, whereas interoperability requires principally technical cooperation and reciprocal confidence in agreed rules. The objective might resemble the architecture of the internet, in which different networks, devices and jurisdictions communicate because they observe common protocols without belonging to a single owner.
Money may eventually need something similar. Preserving a neutral technical layer for international payments would not prevent sanctions or other political restrictions, but it could make wholesale fragmentation of the global financial system less attractive and reduce the danger that rival geopolitical blocs develop entirely incompatible financial infrastructures.
Artificial intelligence and economic sovereignty
The original Bretton Woods institutions were designed for a world of steel mills, railways, currencies and physical reconstruction. Their successors must operate in a world in which computational capacity itself has become a strategic resource. Artificial intelligence complicates geoeconomics because its critical inputs are unusually concentrated: advanced semiconductors, semiconductor-manufacturing equipment, cloud computing, electricity, data and sophisticated human expertise intersect within a comparatively small number of countries and companies.
This creates unprecedented forms of private geopolitical power. A technology company may control infrastructure upon which governments depend, a semiconductor company may occupy a critical point in the military supply chains of dozens of countries and cloud providers may possess technological resources exceeding those available to many states. Economic power that once belonged principally to governments and banks increasingly resides in corporations operating infrastructures whose importance is international.
A new Bretton Woods settlement would therefore need principles governing strategically important private infrastructure. Again, the answer should not be international central planning but resilience. Governments should know whether critical public services depend upon a single foreign technology provider, data should be portable between systems where technically possible and essential governmental infrastructure should possess contingency arrangements. International technical standards should reduce the danger that rival technological ecosystems become permanently incompatible.
The objective would be to preserve the enormous benefits of technological interdependence without permitting interdependence to become technological captivity. This is likely to prove one of the defining economic-policy problems of the next several decades.
Critical minerals without colonialism
Critical minerals provide another test. The green-energy transition, advanced electronics, artificial intelligence infrastructure and modern weapons systems require lithium, cobalt, nickel, copper, graphite, rare earth elements and other materials whose production and processing are geographically concentrated.
This creates an obvious temptation for powerful countries to secure privileged access, and history provides unpleasant precedents. Great powers have repeatedly treated resource-rich weaker countries as territories to be controlled rather than societies entitled to benefit from their own natural wealth. A new competition for strategic minerals could easily reproduce these patterns under the vocabulary of energy transition and national security.
A twenty-first-century Bretton Woods settlement should reject this model explicitly. Development institutions should finance local processing where economically realistic, contracts should be transparent, producing countries should retain meaningful fiscal benefits and infrastructure constructed to extract minerals should, wherever possible, also serve the wider domestic economy.
The goal should be diversified markets rather than exclusive spheres of influence. Africa in particular should not become the terrain upon which China, America and Europe conduct a twenty-first-century version of nineteenth-century resource competition. It should instead become a collection of increasingly prosperous countries able to negotiate with all of them and retain the economic freedom to choose among competing partners.
The right to remain unaligned
This principle points towards a broader political objective. The international economic system should protect the ability of countries not to choose a geopolitical bloc, because the emerging geoeconomic order may otherwise make genuine neutrality increasingly difficult.
During the Cold War many countries attempted to remain non-aligned. The contemporary equivalent becomes harder if technology, finance, trade and security are progressively fused together. If Chinese technological systems become incompatible with American ones, payment networks divide, supply chains are organised exclusively among political allies and investment depends upon security alignment, countries will gradually be forced into economic blocs whether they desire membership or not.
That would be particularly damaging for middle powers and developing countries. A new Bretton Woods system should therefore preserve neutral economic spaces through common payment standards, common commercial-law principles, open shipping routes, multilateral development finance and internationally recognised rules for trade in goods that do not present genuine security concerns.
Countries should remain capable of disagreeing profoundly about geopolitics while continuing to sell one another coffee, machinery or medical equipment. This principle may sound modest, but preserving a substantial sphere of ordinary commerce outside geopolitical competition could prove one of the most important achievements of any revised international economic architecture.
Ukraine and reconstruction
Ukraine would occupy a special place in such a system because her reconstruction will be one of the largest development projects in Europe since the Second World War. The temptation will be to treat reconstruction principally as a financial problem: calculate the damage, raise the money and rebuild what has been destroyed.
That would be a mistake because Ukraine must reconstruct herself for the geopolitical world in which she will actually live. Her electricity system should be decentralised and resilient against attack, her transport infrastructure should connect efficiently with European markets, her defence industry should remain capable of rapid expansion and her digital infrastructure should possess substantial redundancy. Her energy system should minimise vulnerability to hostile external pressure while remaining integrated with the wider European economy.
Reconstruction should therefore not attempt simply to recreate Ukraine as she existed on 23 February 2022. It should build the infrastructure of a country that has learned, at terrible cost, what strategic dependence and infrastructural vulnerability mean.
The same principle applies elsewhere, albeit less dramatically. Infrastructure should increasingly be evaluated not merely according to whether it is efficient in normal circumstances but according to whether it continues functioning when circumstances cease to be normal. Resilience is economically inefficient only if one assumes that emergencies never happen.
No world economic government
There is an obvious danger in everything proposed here. The cure for excessive geoeconomic intervention could become an even larger bureaucracy devoted to managing the world economy, thereby recreating internationally the very planning problem that Hayek identified domestically.
That would be precisely the wrong conclusion. A new Bretton Woods settlement should establish rules, insurance mechanisms and common infrastructure rather than attempt to direct production. Its institutions should normally define the boundaries within which governments and markets operate rather than instruct companies what to manufacture or investors where to place their capital.
The international system cannot know where every factory should be built, but it can establish principles discouraging countries from using subsidies to destroy competitors and subsequently exploiting monopoly positions. It cannot know the correct global production of lithium, but it can encourage transparency and diversified supply. It cannot determine which artificial-intelligence architecture should prevail, but it can encourage interoperability where interoperability is technically feasible.
Institutions should manage the boundaries of competition rather than its outcomes. This distinction is essential if a new international architecture is to reconcile geopolitical realities with economic liberalism rather than quietly abolish the latter in the name of protecting it.
Bretton Woods without Bretton Woods
There may never be another grand conference in a New Hampshire hotel, because the contemporary world is too divided and contains too many centres of power. America cannot design the international economic system largely on her own. China would not accept such an arrangement, India would demand an independent voice, European countries would bring their own interests and the developing world would rightly refuse merely to inherit rules written by others.
The new system may therefore emerge incrementally. Some reforms will occur through the IMF and World Bank, others through the WTO, the Bank for International Settlements and regional development banks. Some arrangements may begin among relatively small groups of countries and subsequently become more universal as their advantages become apparent.
That may actually be beneficial because the twenty-first century does not require another rigid international economic constitution. It needs a network of institutions capable of adaptation as technologies, security threats and economic relationships change. The architecture should therefore be modular rather than monolithic, allowing cooperation to survive even where universal political agreement proves impossible.
Liberalism after innocence
The deepest change required is intellectual rather than institutional. Economic liberalism must lose its innocence without losing its liberalism, because the post-Cold War assumption that commerce could be separated indefinitely from power has proved unsustainable.
Energy can be weaponised, as can finance, technology, shipping and supply chains. Dependence creates efficiencies but also vulnerabilities, and governments cannot reasonably ignore the latter. An economic theory that treats the nationality and political control of every supplier as irrelevant will struggle to survive contact with a world of sanctions, wars and strategic competition.
Yet the opposite conclusion would be equally disastrous. If every economic relationship is interpreted principally through national security, international commerce becomes impossible. Every import becomes a vulnerability, every foreign investment a potential threat, every technology transfer a strategic loss and every trading partner a prospective enemy.
That road leads towards autarky, poverty and ultimately conflict. The task is therefore to preserve a presumption in favour of openness while constructing carefully defined exceptions for resilience. Hayek supplies the presumption, Keynes supplies the exceptions and Bretton Woods supplies the institutional imagination necessary to hold the two together.
A new bargain
The original Bretton Woods settlement rested upon a bargain. Countries accepted international economic rules because those rules left them sufficient domestic political freedom to maintain stable societies, and international liberalism survived because it was not absolute.
The twenty-first century requires a different bargain. Countries should accept continued economic interdependence because international institutions give them reasonable protection against the most dangerous forms of dependency and coercion. In return, they should accept disciplines preventing national security from becoming a universal excuse for protectionism.
Such a bargain would never eliminate geopolitical conflict, nor should anybody imagine that economic institutions could accomplish that. China and the United States will continue to disagree about fundamental questions, Russia will remain a geopolitical problem for Europe for the foreseeable future and Ukraine must organise her economy around the reality of a dangerous neighbour. India will continue seeking strategic autonomy, while middle powers will continue manoeuvring between larger ones.
The objective is not to abolish power politics but to prevent every political dispute from destroying the economic relationships upon which global prosperity depends. That was, in a different form, the achievement of Bretton Woods. Its architects understood that economic institutions cannot abolish political disagreement, but they can prevent political disagreement from producing economic chaos.
Their successors face the same challenge under different conditions. The international economy of the twenty-first century cannot return to the uncomplicated globalisation imagined in the 1990s, but neither should it surrender to a world of fortified economic blocs organised around rival great powers.
Between those alternatives lies a more difficult architecture: open but resilient, competitive but governed by common rules, internationally integrated but conscious of security, technologically interconnected but resistant to coercion. It would recognise that economic efficiency is an indispensable source of prosperity without pretending that efficiency is the only legitimate objective of political communities.
Such an order would be neither Hayek’s world nor Keynes’s. It would contain something of both, because the twenty-first century has rediscovered both the power of markets and the political limits of relying upon them without qualification. That combination may be precisely what a second Bretton Woods requires.
