Essays in distributional econometrics

This dissertation contributes to distributional econometrics by developing new methods to analyze heterogeneity in economic outcomes and applying them to policy-relevant questions. The first chapter proposes a minimum distance estimator for quantile panel data models that is computationally fast, accommodates endogenous group effects, and delivers adaptive inference under non-standard asymptotics. An application to the rollout of the U.S. Food Stamp Program shows that the policy increased birth weight mainly in the lower tail of the distribution. The second chapter develops an econometric framework for studying how treatments affect inequality within and between groups. Outcomes are represented by a two-dimensional quantile surface mapping within-group and between-group ranks to outcome levels. I propose a two-step quantile regression estimator and establish its weak convergence to a bivariate Gaussian process. An application to business training in Kenya shows that treatment effects are concentrated among high-performing firms in strong markets, highlighting complementarities between individual and group performance. The third chapter studies the intergenerational persistence of dietary habits using linked supermarket transaction and administrative data. It documents strong persistence in diet across generations, exceeding income persistence, and shows that only a limited share is explained by the intergenerational transmission of socioeconomic status. Together, the essays show how distributional methods deepen our understanding of inequality, policy effects, and intergenerational transmission.

2025

dissertation

true