Adam Smith Business School

Professor Xavier D’Haultfoeuille (CREST - ENSAE)

Linear Regressions with Combined Data
Friday, 10 October 2025, 15:00–16:30
Room 588AB, Adam Smith Business School

Abstract

We study linear regressions in a context where the outcome of interest and some of the covariates are observed in two different datasets that cannot be matched. Traditional approaches obtain point identification by relying, often implicitly, on exclusion restrictions. We show that without such restrictions, coefficients of interest can still be partially identified, with the sharp bounds taking a simple form. These bounds improve if variables observed in both datasets and not included in the regression are available, even if these variables do not satisfy any restriction. We develop computationally simple and asymptotically normal estimators of the bounds. Finally, we apply our methodology to the estimation of the racial gaps in patent approvals and the effect of patience and risk-taking on educational performance.

Biography

I obtained my PhD at Paris 1 in 2009. After having worked for INSEE, I joined CREST-ENSAE as a professor in economics in 2011. As a researcher, I am mostly interested in (micro-)econometrics, and in particular by treatment effects, (partial) identification and inference issues. Occasionally, I also do some applied work in industrial organization. Since August 2023, I am a managing editor at the Review of Economic Studies.

---

For further information, please contact business-seminar-series@glasgow.ac.uk.

We are committed to providing a collegial, inclusive, and intellectually stimulating environment for all our seminars, in accordance with our established Code of Conduct.

First published: 16 September 2025