Trade, Power, and Economic Coercion in 2025

The Sustainability of Rules-Based Trade

The paper by Cecilia Carvalho, Daniel Monte, and Emanuel Ornelas, models trade regime choice as a dynamic stochastic game where the world’s leader (hegemon or dominant country) selects between a rules-based regime (WTO-style) or power-based regime each period. Leadership transitions follow a Markov process, establishing rules costs while abandoning them is free. The analysis examines Markov-perfect equilibria under unipolar, bipolar, and multipolar leadership structures.

Findings

• Narrow viability window: Rules-based equilibria exist only when the cost to establish rules is moderate, countries are patient, and leadership turnover is high. Too-low costs trigger cycling between regimes while too-high costs lock in power-based systems permanently.
• Inefficiency: Cyclic equilibria waste resources on repeated transitions, while many power-based equilibria persist despite collective preference for rules due to collective-action problems in funding the initial setup cost.
• Bipolarity problem: Two dominant leaders face coordination difficulties, market-power incentives, and free-riding temptations that undermine rules sustainability.
• Multipolarity advantage: Multiple small leaders can more easily sustain rules because market-power incentives are weaker.

How can leaders want to keep rules:

• Expand gains under rules: Make the system more beneficial for everyone so leaders are less tempted by power-based bargaining.
• Exclude a disruptive leader (“WTO–1”): A smaller club can be more stable but loses trade with the big player, so total gains fall.
• “Rules-lite” appeasement: Give leaders extra benefits so they keep the system, but accept lower overall efficiency for the world.

Equilibrium Trade Regimes: Power- vs. Rules-Based
Authors: Cecilia Carvalho, Daniel Monte, Emanuel Ornelas
From: University of Geneva, University of Turin, Sao Paulo School of Economics

Who pressures whom? Mapping global economic statecraft

The paper by Christopher Clayton, Antonio Coppola, Matteo Maggiori, and Jesse Schreger, uses large language models (LLMs) to systematically extract geoeconomic pressure information from hundreds of thousands of earnings-call transcripts and analyst reports (2008–2025).

A two-stage prompt system flags relevant documents first, then processes flagged ones with detailed, instrument-specific prompts to extract: who applies pressure, who receives it, which policy instrument is used, and how firms adjust on prices, quantities, R&D, investment, supply chains, and labor.

Findings

• Geoeconomic pressure is highly concentrated in the hands of the US and China, largely directed at each other. Export controls and sanctions are “small yard, high fence”: narrow sets of sectors and products  such as US semiconductor controls and China rare earths; US/EU sanctions on Russian oil, gas, and finance; targeted sanctions on firms like Huawei.  Recent US tariffs are broad in both sectors and trading partners.
• Using the theoretical “power” index (import dependence × lack of substitutes), the US is significantly more likely to impose export controls on exactly those SIC industries where restricting US supply would hurt China most, consistent with deliberate targeting of chokepoints.
• Tariffs mainly act as a consumer tax in the imposing country: many US firms report higher input costs and raising their own prices. There is little evidence that foreign suppliers cut prices enough to improve US terms of trade. Some firms with more domestic production gain, but on net most report negative effects.
• Export controls on advanced inputs lead targeted foreign firms, especially Chinese tech companies, to invest more in R&D and domestic capacity, and to reconfigure supply chains toward alternative Asian suppliers.
• Sanctions on Russia and Iran hurt targeted entities and some Western firms through lost business and energy-supply risks, while firms in third countries, notably India and China, often benefit via discounted Russian oil or by stepping into vacated trade relationships.
• Responses differ systematically by role: firms in sender countries, receiver countries, and third-party countries show distinct adjustment patterns across prices, quantities, and innovation margins.
• Measurement noise from LLM classification is nontrivial but comparable to well-regarded survey data; cross-model comparisons and human-labeled validation indicate the main stylized facts are robust, though confidence intervals widen once measurement error is explicitly accounted for.

Geoeconomic Pressure
Authors: Christopher Clayton, Antonio Coppola, Matteo Maggiori, Jesse Schreger
From: Yale School of Management, Stanford University, Columbia University