The leaders of the G7 met in Hiroshima this weekend with Ukraine dominating much of the summit. Volodymyr Zelenskyy travelled to Japan, the G7 reaffirmed military, financial and diplomatic support for Ukraine and announced further measures intended to constrain Russia’s war effort. Yet another part of the summit may prove equally important for the structure of the global economy. The G7 formally embraced a strategy of “de-risking, not decoupling” in its economic relationship with China, while promising to diversify supply chains, protect sensitive technologies and coordinate responses to economic coercion.
That phrase represents an important change in how Western governments think about globalisation. For several decades, the dominant assumption was that deeper economic interdependence generally made countries richer and potentially made geopolitical conflict less attractive. Firms located production wherever costs were lowest, supply chains became increasingly international and governments accepted dependence on foreign producers because specialisation increased efficiency. Consumers benefited from cheaper goods, businesses accessed larger markets and countries became economically connected in ways that appeared to raise the price of political confrontation.
Much of that logic remains correct. International trade increases the range of goods countries can consume, allows firms to specialise according to comparative advantage and enables capital and technology to move towards productive uses. Attempting to manufacture everything domestically would make most economies poorer. The G7 therefore explicitly states that its objective is not decoupling from China, and that distinction matters. China is too deeply integrated into global manufacturing and trade for economic separation to occur without enormous costs for both sides.
What has changed is the assumption that all economic dependence should be treated simply as efficient specialisation. Russia’s invasion of Ukraine demonstrated the difference between ordinary interdependence and strategic dependence. Europe benefited for years from relatively cheap Russian energy, particularly natural gas. That relationship lowered costs during normal periods. Once relations collapsed, dependence became leverage because replacing pipelines, suppliers and infrastructure required time and enormous expenditure. The same economic relationship that had previously increased efficiency reduced strategic flexibility during crisis.
The relevant variable is therefore not simply how much countries trade but how substitutable the relationship is. Importing coffee from one country creates relatively little national-security vulnerability if equivalent coffee can quickly be purchased elsewhere. Depending overwhelmingly on one supplier for semiconductors, critical minerals, pharmaceuticals or energy infrastructure creates a different problem because replacement may require years of investment. Concentration becomes strategically important when the product is essential and alternatives cannot be created quickly.
This is why “de-risking” is a more useful concept than complete economic independence. Modern economies cannot eliminate dependence, and trying to do so would destroy many of the gains produced by international specialisation. They can reduce single points of failure. A country importing an essential component from five politically and geographically diverse suppliers remains internationally dependent, but no single government can easily cut off the entire supply.
The concept resembles portfolio diversification in finance. An investor does not reduce risk by refusing to invest anywhere. They reduce risk by avoiding excessive exposure to one asset whose failure would destroy the portfolio. Supply-chain policy can operate similarly. The objective is not autarky but diversification, redundancy and enough spare capacity that one disruption does not disable an entire industry.
The trade-off is cost. Redundancy is almost always less efficient during normal periods. Maintaining alternative suppliers, larger inventories or domestic production capacity can cost more than buying everything from whichever producer currently offers the lowest price. For decades, businesses were rewarded for minimising inventories and constructing extremely lean supply chains. The pandemic demonstrated that systems optimised for ordinary efficiency can become fragile when several links fail simultaneously.
Economic resilience therefore resembles disaster preparedness. A hospital maintaining unused emergency capacity appears inefficient until a pandemic arrives. A government reinforcing buildings appears to spend money on nothing until an earthquake occurs. A company maintaining two suppliers rather than one may pay more every year for a disruption that never happens. The value of resilience appears principally in the counterfactual catastrophe that did not occur.
This creates a political problem because the costs of resilience are visible while the benefits are probabilistic. Consumers can see higher prices today. They cannot observe the future shortage that diversification prevented. Governments therefore need evidence to distinguish genuine strategic vulnerabilities from industries simply using national security as an argument for protection from foreign competition.
That distinction will determine whether de-risking becomes intelligent resilience or ordinary protectionism with new vocabulary. Almost every industry can construct an argument explaining why it is strategically important. Steel matters for defence. Agriculture matters for food security. Pharmaceuticals matter for health. Technology matters for infrastructure. Energy matters for everything. If governments accept every claim, “strategic autonomy” can become a justification for subsidising politically connected domestic industries regardless of economic efficiency.
The correct standard should therefore be relatively narrow: how severe would the national cost be if this supply disappeared, how concentrated is production, and how quickly could substitutes emerge? Semiconductor manufacturing passes that test much more clearly than many ordinary consumer products. Advanced chips are essential to communications, artificial intelligence, vehicles and military systems, while production of the most sophisticated chips is concentrated in a small number of companies and locations. Building equivalent fabrication capacity requires enormous capital, expertise and time.
Critical minerals present a similar problem. Clean-energy technologies require lithium, cobalt, nickel, graphite and rare-earth elements, while processing capacity for several materials is geographically concentrated. The green transition could therefore replace dependence on imported fossil fuels with new dependencies embedded in batteries, electrical systems and renewable technologies unless supply chains diversify alongside decarbonisation.
This is where China becomes central. China is not merely a large exporter of finished goods. It occupies important positions across manufacturing ecosystems, mineral processing, batteries, solar technology and other strategic supply chains. Western economies cannot realistically remove China from these networks quickly, and doing so would substantially increase costs. Yet governments increasingly fear that excessive concentration gives Beijing economic leverage during future geopolitical disputes.
The G7’s answer is therefore deliberately intermediate. Its communiqué states that members seek constructive and stable relations with China while reducing excessive dependencies and countering economic coercion. It also emphasises that a narrow set of technologies with important national-security implications should receive stronger controls without broadly restricting trade.
Whether that distinction remains narrow is the crucial question. Governments usually find it easier to create controls than remove them. A restriction introduced for advanced military-relevant semiconductors can expand towards technologies with increasingly indirect security connections. China can respond with its own restrictions. Businesses then reorganise supply chains in anticipation of further political intervention, making economic separation partially self-fulfilling.
This is the danger of a security dilemma in economics. The United States restricts technology because it fears Chinese military development. China interprets the restriction as evidence that Washington intends to contain its economic rise and increases efforts to achieve technological self-sufficiency. Washington sees those programmes and concludes China is becoming more strategically competitive, producing arguments for further restrictions. Policies initially described as defensive become evidence of hostile intent to the other side.
The same mechanism operates through industrial subsidies. The United States subsidises domestic semiconductor manufacturing because concentration in Asia appears strategically risky. Europe fears investment will move to America and develops its own subsidies. China continues supporting domestic industries. Governments collectively spend more to relocate production according to geopolitical rather than purely economic criteria. Some redundancy may be worth the cost, but the global economy becomes less efficient as security increasingly determines investment.
The question is therefore not whether economic security has a price. It clearly does. The question is which risks are expensive enough to justify paying it. Governments insure houses despite most houses never burning down because the potential loss is catastrophic. They do not insure every possession against every conceivable event because the premiums would exceed the expected benefit. De-risking needs the same discipline.
The Global South complicates the strategy further. Countries such as India, Indonesia, Brazil and Vietnam do not necessarily want to choose between American and Chinese economic systems. China is an enormous trading partner and source of investment, while the United States, Europe and Japan provide other markets, technologies and capital. Their rational strategy may be diversification across both rather than alignment exclusively with either.
The G7 appears increasingly aware of this. Japan invited leaders from India, Indonesia, Brazil, Vietnam, South Korea, Australia, the Comoros and Cook Islands to participate in parts of the Hiroshima summit, while the communiqué emphasised infrastructure partnerships and stronger engagement with developing economies. The geopolitical competition is therefore not simply over military alliances. It is increasingly about who can offer credible economic partnerships.
This matters because Western governments cannot simultaneously ask developing countries to reduce dependence on China while failing to provide alternative infrastructure, investment and market access. China’s Belt and Road Initiative became influential partly because it financed projects governments wanted. Criticising the strategic consequences of Chinese lending is easier than constructing competing financing at comparable scale.
The G7 has reiterated its intention to mobilise up to $600 billion by 2027 through the Partnership for Global Infrastructure and Investment. Whether those commitments become actual projects will matter more than the headline number. Developing countries will judge economic partnerships through roads, ports, energy systems and financing conditions rather than geopolitical speeches.
India is particularly important because its strategy already resembles de-risking without alignment. New Delhi participates in the Quad with the United States, Japan and Australia while remaining in BRICS and the Shanghai Cooperation Organisation. It purchases Russian oil while expanding defence and technology cooperation with Western states. It competes strategically with China while maintaining substantial bilateral trade.
This is not necessarily inconsistency. It is issue-based alignment in a multipolar system. India can cooperate with Washington on Indo-Pacific security, Moscow on energy and defence, Europe on trade and climate, and developing countries on reforming global governance. The ability to maintain several relationships increases bargaining power because no single partner becomes indispensable.
That logic is exactly what the G7 now wants economically. Dependence becomes dangerous when one relationship cannot be replaced. Diversification increases autonomy whether the actor is Germany reducing Russian gas dependence or India refusing to organise its foreign policy around one great-power bloc.
There is an important difference between resilience and economic nationalism, however. Resilience diversifies relationships. Nationalism attempts to relocate everything domestically. The first can preserve much of globalisation while reducing concentrated vulnerability. The second risks fragmenting the world economy into competing blocs with duplicated production and lower productivity.
The G7’s phrase “de-risking, not decoupling” is therefore more than diplomatic language if governments actually maintain the distinction. It recognises that the previous model of globalisation underestimated geopolitical risk without concluding that international economic integration itself was a mistake.
This is probably the more mature lesson from the last several years. The pandemic showed that highly concentrated supply chains can fail because of non-political shocks. Russia showed that economic dependence can be weaponised intentionally. US-China competition shows that technologies once treated as ordinary commercial products can acquire national-security significance.
None of those events disproves comparative advantage.
They demonstrate that efficiency is not the only objective economies optimise.
Resilience has value.
Security has value.
Political autonomy has value.
The difficult part is assigning those values prices rather than treating them as unlimited.
A government willing to pay any economic cost for marginal increases in self-sufficiency will make its citizens unnecessarily poorer. A government unwilling to pay anything for resilience may discover during crisis that the cheapest supplier was expensive in ways its accounting system never measured.
The optimal position lies between them.
That is what Hiroshima’s economic language is attempting to describe. The G7 does not want to stop trading with China, and complete separation would be economically destructive. It wants supply chains sufficiently diversified that economic relationships remain relationships rather than vulnerabilities.
For thirty years, globalisation asked primarily: where can this be produced most efficiently? The emerging system is adding another question: what happens if that producer becomes unavailable tomorrow?
The first question created extraordinary prosperity.
Ignoring the second created fragility.
The next phase of globalisation will depend on whether governments can answer both without sacrificing too much of either.
