PhD Candidate, Department of Economics

Sergey GitlinContact Information

Department of Economics
Northwestern University
2001 Sheridan Road
Evanston, IL 60208

Phone: 847-440-6099

[email protected]

 

 

 

Education

Ph.D., Economics, Northwestern University, 2017 (expected)
MA, Economics, Northwestern University, 2015
Masters, Mathematical Economics and Econometrics, Toulouse School of Economics (Université Toulouse 1 РCapitole), 2011
B.Sc., Economics, University of Nottingham, 2009

Primary Field of Specialization

Econometrics

Curriculum Vitae

Download Vita (PDF)

Job Market Paper

“Efficient estimation with smooth penalization”

Download Job Market Paper (PDF)

This paper proposes an oracle efficient estimator in the context of a sparse linear
model. The estimator optimizes a penalized least squares objective, but, unlike existing
methods, the penalty is once differentiable, and hence the estimator does not engage
in model selection. This feature allows the estimator to reduce bias relative to a
popular oracle efficient method (SCAD) when small, but not insignificant, coefficients
are present. Consequently the estimator delivers a lower realized squared error of
coefficients of interest. Furthermore, the objective function with the proposed penalty
is shown to be convex; paired with differentiability, this ensures good computational
properties of the estimator. These findings are illustrated by simulation evidence, and
the proposed approach is applied to the estimation of effects of location-specific human
capital in agricultural productivity in Indonesia.

Other Research Papers

“Alternative asymptotic analysis of once-differentiable penalty estimator”

Conventional asymptotic analysis of efficient penalized estimators typically prohibits
coefficients of a magnitude that lies in a certain range relative to the sampling error, and it is
well understood that allowing for such coefficients can lead to slower rates of convergence of
such estimators. I derive the asymptotic distribution for the penalized estimator with the once-
differentiable penalty while allowing for coefficients in this range. The analysis is conducted
under standard conditions on the tuning parameters, as well as under an alternative asymptotic
framework that preserves nonlocal properties of these estimators. Inference by a modified
bootstrap procedure is shown to be consistent under an alternative asymptotic view that
excludes intermediate-magnitude coefficients but allows for nonnormal asymptotic distributions
arising from penalization.

References

Prof. Joel Horowitz (Committee Chair)
Prof. Ivan Canay (Committee Co-Chair)
Prof. Alexander Torgovitsky