Semi-Structural Econometric Methods for the Analysis of Inequality

Motivation

Rising income inequality and related disparities, accompanied by high persistence of socioeconomic outcomes across generations, have become a major public policy concern around the world. Documenting inequality, understanding its determinants, and evaluating policies with the goal of reducing inequality is an important part of empirical research in economics.
Widely used statistical tools for quantifying the importance of various drivers of inequality and intergenerational mobility have produced influential findings that are directly, and often informally, interpreted in public and political debates. In these debates, causal and economic meaning tends to be attached to the stylized statistical findings without the use of empirically tested structural models.

Objective

The first objective of this project is to show theoretically and empirically that these widely used statistical tools generate findings that permit a structural economic interpretation only under very strong assumptions. If these assumptions fail, then the structural interpretation may be misleading, and the findings may not in fact measure what they purport to measure.
The second objective is to develop new, robust statistical tools that are guaranteed to produce findings with a structural interpretation while retaining the attractive features of popular existing approaches, e.g., weak assumptions and data requirements.