Alfred Galichon, from NYU, will present his research. 

Optimal transport for economics: applications and extensions

Optimal transport theory has emerged as a unifying geometric framework for a wide range of economic questions that hinge on allocating heterogeneous agents or goods in a cost-efficient way. This talk surveys recent methodological advances and shows how they sharpen three core empirical tools:

- Discrete choice models – I will show how viewing demand estimation through the lens of optimal transport delivers transparent identification of taste heterogeneity and suggests computationally efficient estimation algorithms that scale to high-dimensional product spaces.

- Matching models – Building on the Monge–Kantorovich formulation, I will discuss equilibrium assignments in labor and marriage markets, highlighting new existence and uniqueness results when complementarities are non-transferable. Optimal transport duality provides constructive characterizations of wage/transfer schedules and illuminates comparative-statics of policy interventions.

- Quantile regression and distributional methods – Re-casting quantile regression as an optimal transport problem links conditional quantiles to integral probability metrics. This perspective leads to robust, dependence-aware estimators and facilitates inference on counterfactual distributions, with applications to wage inequality and treatment-effect heterogeneity.

I will show intuition of results, outline computational challenges, and sketch extensions to dynamic settings and strategic interactions, in the hope of showing how optimal transport offers economists a versatile, tractable, and conceptually coherent toolkit for modeling heterogeneity and complementarities across disparate empirical domains.