Time: 12:15 pm – 1:30pm
Date: 3rd July 2019
Place: Room 3001.
Antoine Ferey (CREST) – “Generalized Capital Taxation with Sufficient Statistics” with Benjamin Lockwood (Wharton) and Dmitry Taubinsky (UC Berkeley)
 

Abstract: How should capital and labor income be taxed when people have heterogeneous income-earning abilities and heterogeneous savings preferences? The answer conventionally depends on structural models in which the covariation between income-earning ability and discount rates plays a key role. Yet that covariance is difficult to measure empirically, as are primitives of the structural utility function. This paper uses a novel variational approach to derive a condition for Pareto efficient capital taxes which depends only on the joint distribution of savings and earnings, and on estimable behavioral responses to tax reforms. When heterogeneity is one-dimensional, we show that the optimal income tax takes an intuitive form that parallels the optimal nonlinear labor income tax. We then derive a Pareto efficiency condition in a general setting with multidimensional heterogeneity without any restriction on the tax structure. This characterization of the Pareto efficient system is expressed using empirically estimable statistics that do not depend on social preferences.

Organizers:
Roxana Fernandez Machado (CREST), Marie Laure Allain (CREST), and Linda Schilling (CREST)
Sponsors:
CREST
Lunch registration:
food provided, no registration needed