RESEARCH DOSSIER
Economic Theory under Geopolitical Fragmentation
Abstract. Geopolitical fragmentation changes the feasible organization of economic exchange by conditioning market access, investment, technology transfer, payment settlement, and contractual performance on political relationships and jurisdiction. This monograph examines how the principal mechanisms of economic theory operate under this change. It distinguishes the welfare consequences of restrictions from the potential benefits of reducing exposure to deliberate disruption, and separates geographical relocation from genuine diversification. Comparative advantage, structural gravity, interindustry dependence, investment under uncertainty, financial intermediation, distributional incidence, strategic interaction, and institutional commitment provide complementary explanations of the same process. Evidence from tariff changes, geopolitical shifts in bilateral exchange, commodity-market scenarios, financial stability assessments, and the October 2026 trade outlook is examined according to the distinct estimands and assumptions of each source. Illustrative calculations establish the difference between concentrated exposure and aggregate openness, between the price of redundancy and its conditional insurance value, and between bilateral diversion and global welfare. The analysis concludes that economically defensible resilience requires an explicit account of the particular disruption, the available substitution technology, the allocation of costs, and the institutions responsible for maintaining cooperation. The work is an analytical synthesis of existing economic research; it does not estimate a new structural model or claim a new named theory.
Keywords: economic theory; geoeconomic fragmentation; comparative advantage; structural gravity; trade policy; supply chains; economic security; investment uncertainty; financial fragmentation; welfare; institutional resilience; international cooperation
Introduction
Economic theory studies the coordination of scarce resources among agents whose decisions are interdependent. International exchange expands the set of technologies, factor combinations, suppliers, and buyers available to those agents. Political institutions determine the conditions under which this enlarged opportunity set remains accessible. When governments condition cross-border exchange on strategic affiliation, the opportunity set itself becomes contingent on international political relations. An allocation can continue to satisfy conventional optimizing conditions while being appreciably less productive than an allocation made feasible by a different international institutional arrangement.
This distinction gives the subject its place in economic theory. A tariff modifies a relative price. A prohibition eliminates a choice. An uncertain future licensing condition changes the valuation of an irreversible investment. A payment restriction may make a physically deliverable good commercially inaccessible. Each mechanism has its own comparative statics, time horizon, and incidence. Treating them as a single increase in trade costs conceals information necessary to evaluate their economic effects.
The research question is how a system of decentralized allocation functions when access to other agents and institutions is politically conditional. The relevant unit may be a household, firm, production sector, financial intermediary, jurisdiction, or trading system. Corporate ownership and the settlement infrastructure of a transaction may connect different jurisdictions from the physical route of the merchandise. Therefore, the monograph separates the location of production, the origin of value added, control over assets, and the legal capacity to complete transactions.
The evidence base includes international institutional studies, original economic research, and the conceptual literature of coordination. Observational estimates are treated as estimates of particular relationships within a data vintage; scenario results describe a specified counterfactual; algebraic examples establish logical possibilities. These three kinds of evidence cannot be combined by adding their reported magnitudes. The manuscript evaluates mechanisms rather than attempting an econometric replication.
Historical genealogy and the institutional order
International exchange has repeatedly taken place within changing systems of political authority. The historical account in the WTO's World Trade Report 2026 identifies the collapse of international integration between 1914 and 1945 as a formative episode in the establishment of postwar trade institutions. War, depression, monetary instability, and discriminatory policy formed interacting shocks. The historical sequence illustrates that the international economic order depends on durable expectations about market access and settlement as much as on the availability of transport.
An interwar restriction cannot be understood exclusively as a movement along a demand curve. When firms anticipate foreign retaliation, future market size changes. When financial markets mistrust convertibility, the currency composition of contracts changes. When governments pursue discriminatory arrangements, firms invest in relationships that may become costly to abandon. These feedbacks create a historical path in which the cost of reversing policy can grow with its duration.
The Cold War provides a second comparison. Political affiliation affected the geographical pattern of economic exchange, but the institutional and product structure of today's economy differs. Gopinath and coauthors (2024) use the early Cold War as a comparator in Changing Global Linkages, while identifying third countries that connect present-day blocs. The comparison is valuable as evidence that political alignments can persistently influence exchange. It does not establish the same speed, depth, or duration for current adjustments.
The postwar institutional project codified methods of bargaining and commitment. Reciprocal concessions can internalize some of the cross-border effects of domestic commercial policy. Established procedures can lower uncertainty about what follows a disagreement. Economically, their value includes the investment and adaptation made possible by predictable access. An institution may consequently retain value even during periods when governments disagree about the distribution of the gains from trade.
Historical analogy requires a common measurement basis. Gross goods flows, value-added trade, services, financial claims, and data flows differ in economic meaning and historical coverage. A comparison using nineteenth-century goods trade and contemporary gross exports may overlook the multiple crossings of intermediate goods. History establishes the recurrence of fragmentation and the possibility of institutional reconstruction; quantification requires a separately defined and harmonized outcome.
Definitions and measurement
Geoeconomic fragmentation is a politically induced reorganization or restriction of cross-border economic relationships. Aiyar and coauthors (2023) examine several transmission channels in their study of fragmentation and multilateralism. The economic distinction is between changes in trade caused by evolving costs or preferences and changes attributable to strategically motivated policies or expectations. In practice, causal attribution is difficult because commercial and geopolitical conditions evolve jointly.
Deglobalization denotes a reduction in the relative scale or reach of international integration. Fragmentation concerns the distribution and terms of connections. It can therefore coexist with expanding global trade. Regionalization describes geographical concentration, friend-shoring describes alignment-conditioned relocation, reshoring describes a movement toward domestic production, and diversification describes a reduction in exposure to common sources of failure. These processes may coincide in a particular case; they are analytically distinct.
A useful descriptive starting point is a bilateral trade matrix. Let X(i,j) be the recorded exports from i to j. The fraction of world trade crossing a defined geopolitical partition equals the sum of X across the partition divided by the sum of all international X. Its change must be interpreted alongside the membership rule, commodity composition, valuation basis, and total volume of trade. Changes in nominal commodity prices can alter monetary shares without a corresponding movement in physical quantities.
Consider a hypothetical economy with initial total external trade of 100, of which 40 crosses a geopolitical partition. Later total trade reaches 120, of which 36 crosses it. Aggregate exchange rises by 20%; cross-partition exchange falls by 10%; the cross-partition share moves from 40% to 30%. These quantities describe different aspects of the same observation. The calculation is illustrative and has no external data input.
A country can acquire a different direct supplier while remaining dependent on the same upstream producer. A measure based on customs origin can therefore show diversification when the underlying production network has merely been rerouted. Suitable complementary measures include domestic and foreign value added, ultimate ownership, the jurisdiction of critical production assets, supplier concentration, replacement lead time, settlement currency, and legal restrictions on the use of technology. The appropriate weighting depends on the outcome under investigation.
The October 2026 Global Trade Outlook executive summary reports that the broad divergence between within-bloc and between-bloc trade narrowed during 2025 and into 2026, while US-China decoupling remained an important source of divergence. This later observation places a boundary on extrapolations from earlier estimates. A geographically and sectorally concentrated conflict can coexist with a less pronounced pattern elsewhere.
Comparative advantage and constrained feasibility
Comparative advantage describes gains from specializing according to opportunity costs. Consider two countries A and B and two goods, machinery and food. Suppose a unit of machinery requires two units of labor in A and six in B, while a unit of food requires one in A and two in B. Machinery costs two units of food in A and three in B. A has comparative advantage in machinery and B in food even though A is more productive in both industries. At a terms of trade of 2.5 units of food per machinery unit, each country can access a consumption combination unavailable at its respective domestic opportunity cost.
Now suppose that A's machinery requires a critical license from B. This relationship changes the feasible production set. The labor coefficients are no longer a complete representation of the technology. Where the license is indispensable and unavailable, machinery output may be constrained even when domestic labor and physical equipment are sufficient. If domestic substitution exists but requires an additional investment, the relevant comparison is between accessible technologies over time, including the opportunity cost of installing and operating that substitution.
A tariff can be represented as a wedge between the producer and purchaser price. A quota creates a quantity constraint and may create rents. Export controls can condition eligibility on a product, destination, user, or use. Their economic effects depend on the detailed restriction, enforcement, alternatives, and time horizon. A general equilibrium representation of scarcity must distinguish the cost of using an option from the loss of that option.
A restriction can raise domestic output of an import-competing good while lowering aggregate consumption possibilities. This occurs because resources are moved toward the protected activity despite their alternative productive uses. A strategic benefit may nevertheless justify a cost if a reliable domestic capability reduces a sufficiently valuable public risk. The relevant economic evaluation then includes that benefit, its empirical basis, and the costs of alternative ways to achieve it.
The example also demonstrates why political alignment and economic efficiency cannot be inferred from absolute productivity alone. A country with superior productive capacity in all activities may still benefit from trade according to relative costs. A policy framed as reducing dependence must identify the particular non-substitutable input, rather than treating imports in aggregate as a sufficient measure of vulnerability.
Prices knowledge and allocation
Hayek's 1945 analysis of dispersed knowledge establishes a foundational coordination problem: local information about changing opportunities and constraints must be communicated across agents. Relative prices can summarize information relevant to decentralized decisions. A fragmented price system includes policy wedges that reflect market eligibility as well as production scarcity.
Suppose a component sells for 100 before a restriction and 120 afterward. The difference may contain tariff revenue, changes in freight costs, a scarcity rent, compliance expenses, a risk premium, or supplier markups. A buyer can react to the new price without knowing all of these causes, but a policymaker comparing alternatives must distinguish them. A transfer of tariff revenue within an economy and a real expenditure on redundant documentation have different welfare implications.
Physical shortage and legal inaccessibility can produce the same observed quantity reduction. Where supply is physically scarce, new production can relieve the shortage. Where access is legally constrained, production elsewhere may remain unusable. The choice between building a substitute, increasing inventories, negotiating access, or redesigning a product depends on identifying which constraint binds.
Observed prices also reflect expectations. An immediate price increase can occur if a credible announcement changes future availability and businesses accelerate purchases. Subsequent declines can reflect the exhaustion of stockpiling demand or the arrival of alternative supply. The initial movement is not an unconditional estimate of the long-run cost of fragmentation.
The allocation question consequently extends beyond the informational properties of prices to the institutional definition of an admissible transaction. Price-mediated coordination remains powerful within the feasible set. Membership in that set is partly established by law, contracts, technology, and administrative implementation.
Structural gravity and trade diversion
Structural gravity connects bilateral trade to economic size and relative barriers. Anderson and van Wincoop's Gravity with Gravitas is a foundational contribution to recognizing that trade between two places depends on their barriers relative to alternatives. In a conventional one-sector CES representation, bilateral expenditure can be expressed as
Here Y_i is origin income, E_j is destination expenditure, Y is world income, t_ij is the bilateral iceberg trade cost, the two resistance terms summarize outward and inward market access, and sigma exceeds one. This expression assumes the appropriate CES demand and market-clearing structure. It is presented as a standard theoretical benchmark, not estimated in this monograph.
Holding expenditure, income, and resistance terms fixed, increasing t_ij by 10% with sigma equal to five multiplies bilateral expenditure by 1.1 raised to minus four, approximately 0.683. The partial response is a decline of about 31.7%. In a general equilibrium counterfactual, expenditures and resistance terms respond; the partial response is not the final equilibrium change.
A binary exclusion can be represented by zero admissible trade in a given category, but an aggregate iceberg approximation may hide the different products and firms affected. A small restriction on a specialized input can have an effect far exceeding its direct expenditure share. Structural models must specify the admissible input technologies and product-level margins before their aggregate predictions have the desired interpretation.
Trade diversion occurs when imports from a targeted partner shift toward a different origin. An exempt supplier may gain sales, yet the importing economy may pay more for an equivalent good. Conversely, diversion can preserve access in a disruption and reduce scarcity. Growth in the exempt supplier's exports is insufficient to infer world welfare.
Gopinath and coauthors (2024) report in their bilateral gravity study relative reductions of roughly 12% in bilateral trade value and 20% in the number of announced bilateral foreign direct investment projects between blocs compared with within-bloc flows after Russia's invasion of Ukraine. These are early-period relative model estimates, not changes of those magnitudes in total world trade or the monetary value of investment. In the wider-bloc PPML specification with source-by-time and destination-by-time fixed effects, the trade coefficient (−0.1244; clustered standard error 0.054) is significant at the 5% level, whereas the project-count coefficient (−0.2212; clustered standard error 0.125) is significant at the 10% level. The estimates depend on the bloc definition. The paper also identifies the role of nonaligned connector economies. The estimand, period, and definition of alignment are essential to the interpretation.
Trade responses along existing links and the establishment of new links should be separated. Intensive-margin substitution may use available capacity; extensive-margin entry can require certification, legal contracts, and capital expenditure. Their adjustment speeds differ. If market access uncertainty discourages entry, a measured short-run elasticity may be a poor approximation of the long-run response.
Networks intermediate inputs and bottlenecks
Production is organized through intermediate inputs. A sector that appears small in final consumption may supply a large proportion of other activities. Acemoglu and coauthors' research on aggregate fluctuations shows the relevance of asymmetries in sectoral input-output linkages for the aggregate consequences of sector-specific shocks.
A standard fixed-coefficient input-output demand system is x = Ax + f, where A contains intermediate input coefficients, f is final demand, and x is gross output. If the spectral radius of A is below one, the inverse (I-A)^{-1} exists and x = (I-A)^{-1}f. The inverse captures direct and indirect gross output requirements. This representation describes demand-driven output requirements under fixed coefficients; it does not by itself estimate the effect of a supply prohibition.
For illustration, let A have rows (0.10, 0.40) and (0.20, 0.10), and final demand f have entries 90 and 70. The determinant of I-A is 0.73. The gross output requirements are 109/0.73 = 149.315 and 81/0.73 = 110.959. The spectral radius is approximately 0.383; the fixed-coefficient equilibrium is well defined. If the first sector is unavailable, these requirements cannot be met as stated. A different supply-constrained model, with rationing, inventories or substitution, is needed to estimate actual lost production.
The marginal cost of a component can be small relative to the loss created by its absence. In a Leontief activity producing one final good with one indispensable unit of each of two inputs, the scarce input determines output even if the other is abundant. When substitution is possible, the severity of a shortage depends on available substitutes and the time needed to adapt. A network with more links may be highly redundant or may contain many paths through one common critical node.
Farrell and Newman's 2019 study of weaponized interdependence develops a complementary political account of asymmetric networks. Central nodes can give jurisdictions leverage over information or access. The economic implication is that exposure depends on the location and governance of bottlenecks, as well as the direct commercial relationship between two firms.
A meaningful exposure assessment maps several layers separately: production, logistics, ownership, technology authorization, and payment. A supplier in a politically acceptable country may depend on a technology license or clearing institution in another jurisdiction. Rebuilding one layer while leaving the binding layer unchanged can be costly without improving the desired resilience.
Resilience diversification and uncertainty
Resilience is the ability to maintain or restore specified economic functions after a disruption. Its empirical evaluation requires identifying the function, disturbance, tolerable interruption, and recovery horizon. An economy may be resilient with respect to annual GDP but vulnerable in the continuous supply of a medically or industrially indispensable good.
Diversification is useful when failures are imperfectly correlated. Suppose two suppliers can each meet the customer's entire critical requirement and each has an independent probability of failure of 0.10 during a fixed period. With one supplier the probability of total interruption is 0.10; with both it is 0.01. If their failures are perfectly correlated, the two-supplier interruption probability is again 0.10. These examples assume sufficient spare capacity and immediate substitution. They show why supplier count is an incomplete measure of protection.
Let a one-period interruption produce a real loss of 100 monetary units. A redundant relationship costing four units lowers expected interruption loss from ten to one under independence. Its expected net benefit is five units. With perfect correlation it creates no reduction in expected loss and costs four units. The same procurement decision changes sign when the dependence structure changes. Risk aversion, nonlinear losses, installation delays, and uncertain probabilities would alter the evaluation.
Political shocks can be correlated within alliances, jurisdictions, or infrastructure systems. Domestic production may eliminate one cross-border exposure while increasing exposure to domestic natural disasters, regulatory interruptions, or shortages of imported inputs. A politically allied supplier may be attractive for one disturbance and less attractive for another. Insurance against a named shock must be evaluated against other material vulnerabilities.
The OECD's 2025 Supply Chain Resilience Review reports modeled relocalization scenarios in which global trade declines by more than 18% and real global GDP by more than 5%, with volatility increasing in more than half the economies studied. These outcomes are conditional on the organization's modeled policy experiment, not a forecast of all reshoring programs. They caution against automatically equating domestic location with lower aggregate volatility.
The private decision to hold reserves can differ from the social decision. If a firm's failure creates large downstream losses that the firm cannot fully monetize, its investment in continuity may be too low relative to social welfare. Conversely, publicly financed redundancy may be excessive if beneficiaries select the project but bear little of its opportunity cost. Procurement, insurance, regulation, and public provision have different information and incentive requirements.
Investment technology and dynamic adjustment
Direct investment creates productive relationships that persist beyond the period in which they are established. A firm's decision depends on expected future operating earnings, capital cost, access to intermediate inputs, repatriation of profits, and the probability that legal restrictions change. Sunk costs make a wait-and-see strategy potentially valuable even when the expected future policy wedge is unchanged.
Caldara and coauthors (2019 working paper; 2020 journal version) investigate trade policy uncertainty using several measures of uncertainty and firm-level and aggregate evidence. They associate increased uncertainty with weaker activity and investment and interpret the evidence using a two-country model. This work provides a mechanism through which the announcement and credibility of policy changes can affect allocation before a restriction becomes operative.
A simple investment illustration clarifies the mechanism. An irreversible investment costs 80. If future access remains available, it yields a discounted receipt of 120; if access is lost, it yields 20. With risk-neutral evaluation and probabilities of 0.8 and 0.2 respectively, its expected net present value is 20. With probabilities of 0.5 and 0.5, its expected net present value is minus 10. A real option to delay, if available, has additional value. This calculation describes a one-time decision and omits strategic interaction and financing constraints.
The IMF's April 2023 FDI fragmentation chapter examines the changing geographical alignment of direct investment, emphasizing strategic activities. FDI is economically distinct from portfolio acquisition. A production facility carries organizational knowledge, contracts, managerial routines, and location-specific capital. A financial position can sometimes be rebalanced rapidly, whereas a physical facility can require years to construct, certify and staff.
Adjustment can raise measured investment while reducing its productivity relative to an unrestricted alternative. Rebuilding an existing capability in a new country appears as capital formation. But the resources devoted to duplication could otherwise have expanded capacity or developed new technology. The welfare effect depends on the security value of the new capability, its productivity, and the forgone uses of labor and capital.
International diffusion of technology can depend on imported equipment, worker mobility, joint ventures, intellectual property, access to research, and interoperable standards. Restrictions on one channel may redirect innovation efforts toward indigenous alternatives. The outcome can include both a reduction in the pool of knowledge available to firms and incentives to develop substitute technologies. Which effect dominates is an empirical question by technology, horizon, and institutional arrangement.
Commodity markets and resource constraints
Commodity markets depend on geographically uneven resources, transport routes, storage possibilities, and technical quality requirements. Restricting trade in a commodity with concentrated extraction and low short-run substitutability can divide a previously integrated market into regions with different prices. A surplus in one region can coexist with a shortage elsewhere even when world physical output is unchanged.
Alvarez and coauthors (2023) construct a dataset covering 48 major energy, agricultural, and mineral commodities in their commodity fragmentation analysis. Their partial-equilibrium segmentation scenario identifies vulnerabilities in multiple agricultural markets and minerals important for clean energy, alongside heterogeneous producer and consumer surplus effects. The exercise addresses commodity-market outcomes within its specified structure and does not measure the full general equilibrium cost to the world economy.
A stylized two-market example illustrates the price mechanism. Suppose in an integrated market supply is 2p in each of two regions, with demand 120-p in A and 60-p in B. The integrated equilibrium solves 4p = 180-2p, giving p = 30. A imports 30 and B exports 30. Segmentation yields p_A = 40 and p_B = 20. Under the linear curves and no transport cost, aggregate gains from trade before segmentation equal two triangles of area 150 each, or 300 monetary units. This loss is not the value of eliminated trade; it is the surplus foregone by preventing mutually beneficial marginal transactions.
The model excludes capacity investment and working capital constraints. In the longer run, higher prices may encourage new extraction, alternative technologies, or substitution, while lower prices elsewhere may discourage production. Differences in energy infrastructure and technical specifications can slow that adjustment. Greater supply elasticity lowers the persistence of scarcity, while environmental constraints or lengthy permitting may raise it.
Food and energy prices also redistribute purchasing power between producers and consumers. For a household with limited ability to substitute away from necessities, a price increase changes the feasible budget sharply. For an exporting economy, higher export prices can strengthen external purchasing power while imposing simultaneous costs on domestic users. An aggregate change in GDP can obscure both effects.
Commodity fragmentation connects resource allocation to shared environmental and development problems. Restricting an input critical to a clean technology can raise the cost of a transition. That consequence must be evaluated alongside the security objective and feasible alternatives, rather than treating commodity price changes as the complete social cost.
Money finance and macroeconomic transmission
Cross-border exchange requires means of settlement and financing. A transaction can be admissible under customs law but infeasible under the rules of a relevant financial intermediary. Differences in jurisdiction, currency, settlement arrangements, collateral, and access to foreign-currency liquidity condition the interaction between the real and financial sides of the economy.
The IMF's April 2023 financial fragmentation chapter identifies capital allocation, payment infrastructure, bank funding, lending and risk-sharing channels. A rise in geopolitical risk can produce a reassessment of the probability of cross-border claims being impaired or inaccessible. A financial intermediary exposed to the reassessment may reduce credit to its domestic borrowers even when the direct trade channel is initially limited.
Balance-sheet effects are central. A bank with foreign-currency liabilities and domestic-currency assets can experience a deterioration in its net position after a depreciation. A firm with matched currency assets and liabilities faces a different direct currency risk but may still face refinancing and settlement risk. The relevant measure is the entire contractual cash flow and maturity structure, rather than gross external debt alone.
External risk sharing depends on the correlation of consumption or income shocks and the enforceability of claims. A reduction in international asset access can lower exposure to an adversarial jurisdiction and simultaneously reduce diversification against domestic disturbances. The optimal portfolio under a credible threat of exclusion differs from the optimal portfolio under ordinary credit risk.
Macroeconomic adjustment depends on price and wage flexibility, exchange-rate regimes, external financing, and fiscal space. Consider an import cost increase that reduces productive capacity and raises the domestic price level. Monetary accommodation may cushion nominal demand while intensifying inflationary pressure; tighter policy may stabilize expectations while amplifying the output contraction. One-time relative-price adjustments should be distinguished from sustained inflation, which depends on propagation through wages, expectations, demand and policy.
A static accounting identity illustrates why trade balances alone are not sufficient. National saving minus domestic investment equals the current account under the usual accounting definitions. A bilateral goods deficit can diminish while the overall current account is broadly unchanged, as transactions move to other partners or domestic saving and investment adjust. A trade restriction changes the economic determinants of the identity, but the identity alone cannot predict their response.
Fiscal policy can redistribute losses, finance alternative infrastructure, or stabilize incomes during transition. Funding has opportunity costs and may shift burdens to future taxpayers. For countries facing external financing constraints, the fiscal cost of a supply strategy must be evaluated with its impact on currency needs and debt sustainability. A fiscally affordable program and a physically feasible project need not coincide.
Distribution welfare and the second best
Economic welfare analysis distinguishes transfers from real resource costs, consumer and producer effects, and public benefits. A tariff may raise domestic producer surplus and government receipts while reducing importer and consumer surplus. The allocation of the burden depends on supply and demand elasticities, market power, currency adjustment, and enforcement.
Amiti, Redding and Weinstein's (2019) study of the 2018 tariffs estimates substantial pass-through to tariff-inclusive import prices and a deadweight welfare loss of $1.4 billion per month by the end of 2018. The magnitude is a period-specific monthly estimate. It does not establish that exporters always bear zero incidence, or that this monthly flow should be compared directly with an annual national income measure.
Fajgelbaum and coauthors (2020) distinguish gross losses to US import buyers, estimated at $51 billion or 0.27% of GDP, from aggregate real-income losses, estimated at $7.2 billion or 0.04% of GDP after accounting for government revenue and domestic producer gains, in The Return to Protectionism. These model-based quantities refer to the studied episode and illustrate why gross distributional burdens are different from net welfare changes. They should not be added to the preceding estimate: the studies examine overlapping policies using different welfare calculations and reporting periods.
The following welfare distinction is general. Suppose a policy creates a domestic transfer of six units from buyers to the government and causes a real efficiency loss of two units. For an equally weighted domestic aggregate, the transfer cancels in consolidation; the real loss remains. For a household with little disposable income, losing six units may be important even if a government accounting entry offsets the transfer at the aggregate level. Social welfare weights, compensation, and the actual use of revenue determine how the aggregate should be interpreted.
A security externality can change the welfare conclusion. If a privately efficient input source creates a social exposure that market participants do not internalize, interventions might improve welfare. The case for intervention depends on a demonstrated externality, the incremental reduction of risk, fiscal and administrative costs, retaliation, and alternative instruments. A targeted reserve or continuity requirement may address a bottleneck with fewer distortions than a broad import restriction. There is no universal ordering across these instruments without information about the shock and the feasible policy set.
Second-best reasoning matters when multiple distortions coexist. Removing a single policy wedge does not necessarily restore the first-best allocation if market power, unpriced risks, contractual incompleteness, or another restriction remains. Similarly, the existence of one distortion does not imply that introducing any further distortion improves welfare. The comparison is between specified feasible alternatives, using a consistent welfare criterion.
Distribution has a political consequence. Regions with concentrated import competition and regions exposed to retaliatory restrictions may favor different policies. Even if aggregate real income increases from exchange, adjustment costs can be large for workers with occupation-specific human capital. An institution capable of supporting mobility, income insurance, or retraining can change the distribution and hence the political sustainability of integration.
Strategic interaction and institutional resilience
Political authorities interact strategically. A tariff chosen in isolation may alter foreign export prices or domestic sectoral rents. When a partner responds, the resulting policy may leave both economies with a lower level of welfare than a cooperative arrangement. This is a familiar coordination problem rather than a novel characterization of international relations.
Consider two symmetric countries with illustrative payoffs expressed as welfare units. Mutual open access gives each 8. One-sided restriction gives the restricter 9 and the open partner 4. Mutual restriction yields 5 each. Restriction is individually attractive against either response, while mutual openness yields superior joint welfare. In repeated interaction, credible future responses and cooperative institutions can change incentives. The numbers are stipulated, not calibrated; reversing the ranking or adding a genuine security benefit can change the strategic outcome.
The central institutional problem is credibility. Governments cannot simply declare a future restriction impossible if subsequent political authorities retain discretion. A binding commitment, transparency requirement, dispute procedure, or well-specified contingency clause can make future market access more predictable while preserving legitimate emergency capacity. Enforcement architecture influences investment incentives because firms care about the expected durability of the commitment.
The WTO's 2023 re-globalization report argues for broadening participation and international cooperation as a response to shared security, development, and environmental challenges. Its institutional position is relevant to the economic comparison of multilateral and fragmented arrangements. A successful institutional arrangement must address the actual distribution of risks and benefits, not rely on aggregate gains alone.
Security objectives can conflict across jurisdictions. An action that reduces one country's perceived vulnerability may increase another's, prompting reciprocal restrictions. Credible cooperation can be easier where parties agree on the treatment of goods with substantial civilian externalities or can verify compliance with mutually understood rules. Where objectives conflict fundamentally, economic analysis can still illuminate the costs and feasible margins of narrower arrangements.
Institutional resilience is the capacity to preserve economically useful commitments and to manage disagreement through predictable processes. It requires information, administrative capability, financing, credible dispute resolution, and mechanisms for adapting obligations to new conditions. An arrangement that cannot adjust may lose legitimacy; one with unconstrained discretionary exceptions may lose credibility. The challenge is to maintain the economic value of commitments under changing political constraints.
Comparative evidence and empirical identification
The available research measures distinct aspects of fragmentation. An episode-based tariff study uses variation in product exposure and prices to assess incidence. A bilateral gravity study examines changes in trade between groups relative to a comparator. A commodity simulation specifies a counterfactual segmentation and evaluates price and surplus effects. An economy-wide scenario describes a modeled equilibrium under specified restrictions. Each approach can illuminate a mechanism; their estimates do not measure one interchangeable parameter.
| Study or vintage | Object | Interpretation |
|---|---|---|
| Amiti, Redding and Weinstein, 2019 | US tariffs during 2018 | Price incidence and end-2018 monthly real-income effect |
| Fajgelbaum and coauthors, 2020 | US policy and retaliation in 2018 | Gross buyer loss versus consolidated model-based real-income loss |
| Gopinath and coauthors, 2024 | Post-invasion bilateral flows | Relative cross-bloc changes in trade value and announced FDI project counts within the early-period specification |
| Alvarez and coauthors, 2023 | Segmented commodity markets | Partial-equilibrium surplus and price effects under hypothetical grouping |
| OECD, 2025 | Relocalization experiment | Conditional global trade, real GDP and volatility effects |
| WTO, October 2026 | World goods trade and alignment | Updated descriptive pattern and dated forecast |
The bibliography identifies the particular versions supporting these claims. For the last row, the WTO October 8 forecast announcement gives the revised 2026 world merchandise trade volume growth forecast of 3.9%, up from the March baseline projection of 1.9%. This is a change in forecast vintage, not a realized full-year growth rate or a direct causal estimate of fragmentation. The revision is compatible with ongoing changes in particular trade relationships.
The principal identification challenge is joint determination. Political tension can respond to trade dependence; trade restrictions can respond to adverse economic developments; investment can move in anticipation of both. Static correlations between policy and flows therefore do not determine the causal effect of policy.
A conventional empirical design would use data at origin-destination-product-time level, a clearly dated restriction, and pre-policy exposure measures. Origin-time and destination-time effects can absorb broad changing economic conditions; pair fixed effects can absorb persistent bilateral differences; sector-time effects can absorb common industry shocks where the data permit. These controls depend on the measurement level and the variation which remains. If the restriction operates across every destination at the same time for the same origin, fixed effects may absorb part or all of the intended treatment.
A gravity estimate in levels using Poisson pseudo-maximum likelihood can accommodate zero flows and many forms of heteroskedasticity under appropriate conditional mean assumptions. A log specification drops zeros unless a defined transformation or alternative model is used. Neither estimator guarantees causal identification. Anticipation, heterogeneous treatment timing, spillovers into comparator countries, and simultaneous shocks require separate attention. Standard errors must account for the level of policy variation and dependence.
A difference-in-differences design requires a plausible counterfactual evolution of treated outcomes in the absence of treatment. Pre-trend tests are informative but cannot prove this assumption. A supplier-country control may itself be affected by diversion; that contamination changes the meaning of the estimate. When a worldwide shock has network effects, an exposure design or explicit general equilibrium counterfactual may be more suitable than a simple treated-versus-untreated comparison.
Measurement robustness should compare alternative definitions of geopolitical alignment, real quantities versus nominal values, and direct shipment origin versus value added. Foreign direct investment announcements, realized flows, positions, and ultimate ownership measure different stages of capital allocation. Comparing those measures can reveal whether a change reflects intended future investment or actual reorganization of capacity.
A future replication would obtain the original data, code, fixed effects, sample definitions, confidence intervals, and alternative scenarios for each quantitative study. This monograph does not reproduce their estimation results. Its empirical contribution is a structured comparison of the meaning and scope of the reported findings, supported by the accessible research records and institutional summaries.
Original analytical synthesis
The research supports a distinction among the value of connectedness, the expected consequences of disruption, and the value of the institutions governing access. The first concerns opportunities for specialization, exchange, and knowledge transfer. The second concerns the state-contingent availability of these opportunities. The third concerns the credibility with which access is maintained or modified. Conventional welfare analysis can incorporate all three when the relevant choices, states of nature, and social objectives are specified.
Three analytical implications follow from the mechanisms developed above.
First, a change in the organization of international links can have a large effect on welfare without producing a commensurate change in aggregate trade volume. Consider the substitution of a more expensive supplier after the elimination of a cheaper source. The volume can remain stable while the consumption possibilities supported by that volume decline. Conversely, substitution may avert a much larger loss than the increase in procurement cost. Aggregate trade growth is consequently compatible with either a high or a low welfare cost of restricted access.
Second, the economic cost of a security strategy must be evaluated relative to the risk it actually reduces and the alternatives available. The redundancy example demonstrates that the insurance value of a second supplier depends on the joint probability of interruption, while the input-output example demonstrates why a small missing input can constrain large outputs. Combining these insights, a broad measure of import exposure is an inadequate ranking of security priorities. A cost-effective response identifies the function at risk, the shared causes of failure, the time required for substitution, and the downstream effects. This is an application of established decision and production theory, not a newly proposed theory.
Third, credibility has economic value before the physical quantity of exchange changes. When an investment is irreversible, policy uncertainty can alter entry and capital allocation in advance of actual restrictions. The expected quality of institutional commitments therefore affects productive capacity. Declines in the durability of commitments can have effects that are difficult to observe in contemporaneous trade statistics. A recovery of flows does not necessarily establish that long-horizon commercial confidence has fully recovered.
These conclusions yield a practical distinction between a national welfare assessment and a world welfare assessment. A country's gain from diverting trade, collecting policy rents, or securing access may coincide with a loss elsewhere. A world evaluation consolidates transfers among jurisdictions, while a national evaluation may treat them as part of its own income. Public safety and security objectives may be valued differently by governments. Any aggregate estimate must state the welfare boundary and normative assumptions.
The economic choice is also intertemporal. Maintaining duplicate production can sacrifice current output to preserve access during a future state. A calculation must include the probability and duration of that state, the reliability and cost of alternatives, and the timing of investments. A zero-disruption baseline provides one counterfactual; a realistic disruption baseline can provide another. Reporting both makes explicit the security benefit that would otherwise remain implicit.
International institutional resilience and domestic risk management are therefore complementary lines of analysis. Greater domestic continuity can be valuable when international commitments cannot be trusted. Predictable international commitments can reduce the need to duplicate costly fixed assets. Diversified sourcing can improve continuity without sacrificing the advantages of specialization where dependencies are sufficiently independent. These strategies address different failure mechanisms and can be evaluated within the same resource and welfare accounting.
For economic theory, the principal interpretive result is that geopolitical restrictions act on the institutional conditions of feasible exchange. Modeling them requires separate representations of prices, quantities, permissions, expectations, and network relations where the application warrants the complexity. A simple tariff-equivalent approximation can be useful for some questions and inadequate for others. The choice of representation follows the objective of the analysis, not the political label attached to the restriction.
Limitations and open research problems
The evidence reviewed here is selective and methodologically heterogeneous. The manuscript uses original-text reproductions, journal or research-institution abstracts, official chapter summaries and indexed source excerpts. Some full-text endpoints returned access errors during this research session. Quantitative claims have been confined to results explicitly available in the cited records. A comprehensive full-text audit of methods, statistical uncertainty, data preparation and scenario sensitivity has not been completed. This matters particularly for interpreting the gravity and economy-wide simulation estimates.
The formal illustrations use stipulated coefficients and probabilities. They are transparent examples rather than calibrated descriptions of specific countries. The input-output exercise is a demand-driven calculation; no supply-side multiplier is inferred from it. The strategic game omits asymmetric power and repeated-game enforcement. The commodity model excludes transport costs, inventory, uncertainty and endogenous capacity. The monetary discussion describes mechanisms rather than estimating their magnitude.
A complete empirical assessment of fragmentation should establish a common time series of trade, foreign value added, ultimate investment ownership, payment relationships, and product-specific substitution. Consistency across data sources remains a substantive research problem. Policies that target goods by user or use can also be challenging to represent in conventional customs datasets.
Future research should examine the joint distribution of supplier failures rather than infer independence from the number of suppliers. It should measure substitution lead times and downstream production losses, distinguish early diversion from long-term capital formation, and compare the observed cost of redundant facilities with counterfactual disruption losses. Identification of precautionary investment requires estimating the behavior that would have occurred absent the perceived threat.
The analysis of institutional effectiveness similarly requires measurable outcomes. Market access commitments, the predictability of enforcement, payments continuity, and the time required to resolve a disagreement are candidate measurements. Their causal effects on investment and productivity are difficult to separate from state capacity and macroeconomic stability.
The October 2026 institutional evidence is a dated snapshot. The forecast revision and aggregate alignment pattern are not projections made by this monograph. No source is used to claim that geopolitical fragmentation follows an irreversible trajectory. New evidence may change the magnitude and direction of particular relationships, while the allocation and coordination mechanisms remain analytically relevant.
Conclusion
Geopolitical fragmentation reorganizes the feasible conditions of exchange. Its economic consequences arise through changes in relative prices, eligibility, production networks, investment decisions, financial claims and institutional commitments. Comparative advantage continues to explain the opportunity costs of restricted specialization; structural gravity clarifies why relative access matters; network economics explains why small but critical inputs can produce disproportionately large disruptions. Decision theory and welfare economics establish how continuity benefits can be weighed against resource costs.
The historical record illustrates that political restrictions can transform international integration over extended periods. Contemporary evidence establishes substantial changes in specific bilateral flows and tariffs, while later trade observations demonstrate that aggregate expansion and partial reversal of bloc divergence can occur alongside persistent strategic conflict. These observations require distinct time horizons and estimands.
Resilience is economically valuable when it preserves a specified function under a plausible interruption. Diversification depends on joint exposure, redundancy has a resource cost, and domestic location protects against some but not all disturbances. The distribution of costs and the credibility of institutions determine whether particular responses can endure. The most informative economic assessment specifies the feasible alternatives, quantifies the particular risk where evidence permits, and reports effects on households, producers, public budgets and external partners.
The central conclusion is that international economic theory remains applicable under geopolitical fragmentation when it makes institutional access, uncertainty, and interdependence explicit. The task of analysis is to reveal the consequences of available arrangements and their trade-offs with precision appropriate to the evidence. No general political mandate follows from a single aggregate estimate.
References
- Acemoglu, Daron, Vasco M. Carvalho, Asuman Ozdaglar, and Alireza Tahbaz-Salehi. 2012. The Network Origins of Aggregate Fluctuations. Econometrica 80(5): 1977-2016. Universitat Pompeu Fabra working paper 1291 (2011), research record. Source
- Aiyar, Shekhar, Jiaqian Chen, Christian H. Ebeke, Roberto Garcia-Saltos, Tryggvi Gudmundsson, Anna Ilyina, Alvar Kangur, Tansaya Kunaratskul, Sergio L. Rodriguez, Michele Ruta, Tatjana Schulze, Gabriel Soderberg, and Juan P. Trevino. 2023. Geoeconomic Fragmentation and the Future of Multilateralism. IMF Staff Discussion Note 2023/001. DOI: 10.5089/9798400229046.006. Source
- Alvarez, Jorge A., Mehdi Benatiya Andaloussi, Chiara Maggi, Alexandre Sollaci, Martin Stuermer, and Petia Topalova. 2023. Geoeconomic Fragmentation and Commodity Markets. IMF Working Paper 2023/201. DOI: 10.5089/9798400252426.001. Source
- Amiti, Mary, Stephen J. Redding, and David E. Weinstein. 2019. The Impact of the 2018 Tariffs on Prices and Welfare. Journal of Economic Perspectives 33(4): 187-210. DOI: 10.1257/jep.33.4.187. Source
- Anderson, James E., and Eric van Wincoop. 2003. Gravity with Gravitas: A Solution to the Border Puzzle. American Economic Review 93(1): 170-192. DOI: 10.1257/000282803321455214. Earlier NBER Working Paper 8079 (2001), DOI: 10.3386/w8079. Source
- Caldara, Dario, Matteo Iacoviello, Patrick Molligo, Andrea Prestipino, and Andrea Raffo. 2019. The Economic Effects of Trade Policy Uncertainty. International Finance Discussion Papers 1256. Board of Governors of the Federal Reserve System. DOI: 10.17016/IFDP.2019.1256. Published journal version, 2020, Journal of Monetary Economics 109: 38-59, DOI: 10.1016/j.jmoneco.2019.11.002. Source
- Fajgelbaum, Pablo D., Pinelopi K. Goldberg, Patrick J. Kennedy, and Amit K. Khandelwal. 2020. The Return to Protectionism. Quarterly Journal of Economics 135(1): 1-55. DOI: 10.1093/qje/qjz036. Online 2019; Yale research record. Source
- Farrell, Henry, and Abraham L. Newman. 2019. Weaponized Interdependence: How Global Economic Networks Shape State Coercion. International Security 44(1): 42-79. DOI: 10.1162/ISEC_a_00351. Belfer Center research record. Source
- Gopinath, Gita, Pierre-Olivier Gourinchas, Andrea F. Presbitero, and Petia Topalova. 2024. Changing Global Linkages: A New Cold War? IMF Working Paper 2024/076. DOI: 10.5089/9798400272745.001. Source
- Hayek, Friedrich A. 1945. The Use of Knowledge in Society. American Economic Review 35(4): 519-530. Econlib reproduction. Source
- International Monetary Fund. 2023. Geoeconomic Fragmentation and Foreign Direct Investment. Chapter 4 in World Economic Outlook: A Rocky Recovery, April 2023. Washington, DC: IMF. Source
- International Monetary Fund. 2023. Geopolitics and Financial Fragmentation: Implications for Macro-Financial Stability. Chapter 3 in Global Financial Stability Report: Safeguarding Financial Stability amid High Inflation and Geopolitical Risks, April 2023. Washington, DC: IMF. Source
- OECD. 2025. OECD Supply Chain Resilience Review: Navigating Risks. Paris: OECD Publishing. DOI: 10.1787/94e3a8ea-en. Source
- World Trade Organization. 2023. World Trade Report 2023: Re-globalization for a Secure, Inclusive and Sustainable Future. Geneva: WTO. Source
- World Trade Organization. 2026. AI Boom and Trade Resilience to Middle East Crisis Lift Outlook, but Gains to Be Uneven. 8 October 2026. Geneva: WTO. Source
- World Trade Organization. 2026. Global Trade Outlook and Statistics: Update, October 2026. Geneva: WTO. Executive summary. Source
- World Trade Organization. 2026. World Trade Report 2026. Geneva: WTO. Official report overview and chapter summaries. Source