I am an Assistant Professor of Economics (Econometrics) at the University of Michigan.

My research interests are nonparametric approaches for statistical identification, estimation, and inference. My current focus is on developing geometric methods for causal inference and statistical optimal transport theory.



ffg [at] umich [dot] edu

Research papers

6. Distributional synthetic controls (major revision requested, slides)

5. A path-sampling method to partially identify causal effects in instrumental variable models (submitted, slides)

This paper by Kilbertus, Kusner and Silva (2020) introduces a more efficient and robust (albeit less general in that no general assumptions on individual behavior can be included in the optimization) practical estimation procedure by reducing the problem to a classical GMM problem. It also contains an additional analysis of my general implementation.

4. Non-testability of instrument validity under continuous treatments (forthcoming at Biometrika)

3. On the convergence rate of potentials of Brenier maps (accepted at Econometric Theory)

2. A condition for the identification of multivariate models with binary instruments (revision requested)

  1. Independent nonlinear component analysis (joint with Susanne Schennach, submitted)