Quantitative Trade Models

What quantitative trade models do

A quantitative trade model takes a general-equilibrium theory of trade, disciplines it with data on bilateral flows, tariffs, and production, and uses the calibrated structure to evaluate counterfactual policies. The workhorse framework descends from Armington and Eaton-Kortum: countries differ in productivity, goods are differentiated by origin, and trade costs drive a wedge between production and consumption patterns. Gravity estimation identifies the key elasticities; the calibrated model then answers questions like "what happens to welfare if tariffs rise by 10 percentage points?" or "what is the optimal tariff schedule?"

The papers linked here push this framework in several directions.

Nesting and generality

The WTO dissolution paper builds a non-parametric neoclassical model that nests a wide class of quantitative trade models. It derives analytic formulas for optimal tariffs that depend on observable sufficient statistics rather than on a particular parametric assumption about demand or supply. The practical value is that the formulas can be applied across model variants without re-solving each one, and they deliver the $2.8 trillion dissolution-cost estimate that anchors the trade agreements hub.

Discrete industries and non-traded goods

Standard models treat the number of traded varieties as continuous, which works well for differentiated manufactures but poorly for industries where entry is lumpy. Discrete Trade generalizes the framework to accommodate both discrete and continuous industries. The model introduces affordability constraints and non-traded service prices as determinants of the trade pattern, producing gains-from-trade estimates that differ from continuous benchmarks and generating big-push effects where small policy changes trigger discrete entry.

Technology adoption and distortions

The trade and technology adoption paper embeds firm-level technology choice and labor-market distortions into a quantitative trade model. The central finding is that distortions erode roughly one-third of the productivity gains from trade liberalization. The mechanism runs through misallocation: when labor markets are distorted, firms adopt technologies that are privately optimal but socially inefficient, and trade amplifies those choices. This matters for gains-from-trade calculations because it means that standard models, which assume undistorted factor markets, overstate the welfare benefits of openness in economies where distortions are large.

Trade costs and quality

Weight-based quality specialization validates the iceberg trade-cost assumption by showing that quality specialization occurs along the weight margin: heavier goods bear higher per-unit shipping costs, so distant exporters specialize in lighter, higher-quality varieties. The paper provides micro-level evidence that the functional form used in nearly every quantitative trade model is not merely convenient but empirically grounded.