Professor Seiler’s research focuses on the intersection of machine learning and consumer search behavior, utilizing large-scale data to understand decision-making across diverse sectors—from consumer goods to healthcare. His recent work addresses a fundamental challenge in causal inference: estimating demand when prices adjust endogenously in response to a treatment.
By applying his methodology to supermarket scanner data, Professor Seiler demonstrates how traditional regressions can lead to substantial bias when prices are not properly instrumented. His research provides a robust framework for testing and recovering unbiased treatment effects, offering critical methodological guidance for scholars and practitioners analyzing the impact of advertising and pricing strategies
