Sanction-Driven Trade Diffusion and Substitution: A Gaussian Process Model with Structured Latent Confounders

• In Event: Lightning Talk: Measurement, Modeling, and Political Behavior

Brief Overview

This paper introduces Gaussian Process Modeling with Structured Latent Confounders (GPM-SLC) to study sanctions. Using Russia’s 2014–2023 trade, it shows how sanctions shocks reshape networks through diffusion, substitution, and avoidance.

Author

• Jason Berteotti, Purdue University