Quantitative analysis of income and wealth distributions using novel econometric methods

We will quantify the relevance of different sources of inequality at different parts of the income (or wealth) distribution.

We consider the following sources of inequality: gender-based wage discrimination, family background, marital sorting and endogenous labor market participation.The standard approach to quantify the relevance of these sources is to estimate linear models for the conditional mean of income (or wealth).

This approach focuses on average effects but does not reveal heterogenous effects along the income (or wealth) distribution.

Average effects are uninformative about e.g. the impact of inequality at the tails of the income or wealth distribution, including inter-generational poverty traps or the “glass ceiling” on female wages.We will conduct a full distributional analysis and will be able to quantify how these different sources of inequality affect the shape of observed income or wealth distributions.

This is econometrically challenging: important features of the labor market, including endogenous selection into employment and relevance of unobserved characteristics (fixed effects) cannot be accounted for by using standard tools for distributional analysis.

We plan to use modern econometric methods that so far have not been applied in this empirical context, some of which we will extend or develop as part of this project. These methods are based on distributional regression, counterfactual distributions, and deconvolution techniques.