Quantitative Trade Models

What quantitative trade models do

A quantitative trade model combines a general-equilibrium theory with data on bilateral trade, tariffs, and production, then uses the calibrated equilibrium to evaluate counterfactual policies. In the workhorse frameworks associated with Armington and Eaton-Kortum, countries differ in productivity, goods are differentiated by origin, and trade costs create a wedge between production and consumption. Gravity estimation identifies key elasticities used to compute, for example, the welfare effect when tariffs rise by 10 percentage points and the optimal tariff schedule.

Nesting and generality

The WTO dissolution paper develops a non-parametric neoclassical model that nests a broad class of quantitative trade models. Its analytic optimal-tariff formulas depend on observable sufficient statistics rather than a particular parametric specification of demand or supply. As a result, the formulas apply across model variants without requiring each one to be solved anew. They also deliver the $2.8 trillion estimate of the cost of WTO dissolution reported in the trade agreements hub.

Discrete industries and non-traded goods

Standard models treat the number of traded varieties as continuous, an approximation that is less suitable when entry occurs in discrete units. Discrete Trade allows for both discrete and continuous industries. Affordability constraints and the prices of non-traded services help determine trade patterns, yielding gains-from-trade estimates that differ from continuous benchmarks. The model also generates big-push effects in which a small policy change induces discrete entry.

Technology adoption and distortions

The trade and technology adoption paper adds firm-level technology choice and labor-market distortions to a quantitative trade model. The model implies that these distortions erode roughly one-third of the productivity gains from trade liberalization. The mechanism is misallocation: firms facing distorted labor markets adopt technologies that are privately optimal but socially inefficient, and trade amplifies these choices. Standard gains-from-trade calculations, which assume undistorted factor markets, therefore overstate the benefits of openness in economies with large distortions.

Trade costs and quality

Weight-based quality specialization evaluates the iceberg trade-cost assumption using specialization along the weight margin. Heavier goods face higher per-unit shipping costs, so distant exporters specialize in lighter, higher-quality varieties. This micro-level evidence supports the trade-cost specification used in nearly every quantitative trade model.