Accepted and Published papers

• Dufays, A., Houndetoungan, E. A., & Coën, A. (2022). Selective linear segmentation for detecting relevant parameter changes. Journal of Financial Econometrics, 20(4), 762-805.

  Paper WP

Working Papers

• Identifying Peer Effects on Student Academic Effort, 2023 (with Cristelle Kouame)

  Peer influences on academic outcomes are often studied using GPA as a proxy variable for students' academic effort. We show that this may bias the peer effects when the network includes isolated students. This is because GPA depends on unobserved factors that do not influence isolated and non-isolated students in the same way. To resolve this issue, we propose an identification strategy that distinguishes unobserved shocks exerted on GPA without influencing the effort from unobserved students' preference shocks. Applying our approach to Add Health data shows that the peer effect estimate using standard approaches is biased downward.

  Paper Online Appendix Codes (R & C++)

• Estimating Peer Effects using Partial Network Data, 2023 (with Vincent Boucher)

  We study the estimation of peer effects through social networks when researchers do not observe the entire network structure. Special cases include sampled networks, censored networks, misclassified links, and aggregated relational data. We assume that researchers can obtain a consistent estimator of the distribution of the network. We show that this assumption is sufficient for estimating peer effects using a linear-in-means model. We provide an empirical application to the study of peer effects on students academic achievement using the widely used Add Health database and show that network data errors have a first-order downward bias on estimated peer effects.

  Paper Online Appendix R Package Vignette

• Count Data Models with Social Interactions under Rational Expectations, 2022 (Reject and Resubmit at the Journal of Econometrics)

  I propose a peer effect model for counting variables using a game of incomplete information. I present sufficient conditions under which the game equilibrium is unique and study the model parameter identification. My identification result is general and can be applied to other nonlinear models with social interactions, such as binary and ordered response models. I show that the parameters can be estimated using the Nested Partial Likelihood (NLP). I generalize the estimator to the case of endogenous networks and study its asymptotic properties. I show that the linear-in-means/Tobit models with a counting outcome are particular cases of my model. However, these models ignore the counting nature of the dependent variable and lead to inconsistent estimators. I use the model to evaluate peer effects on students' participation in extracurricular activities. I find that increasing the expected number of activities in which a student's friends are enrolled by one implies an increase in the expected number of activities in which the student is enrolled by 0.08. The estimate of this effect with the Tobit model is three times higher.

  Paper Online Supplement R Package

Work In Progress

• Inference for Sequential Conditional M-estimators (with Abdoul Haki Maoude)

  We propose a simulation-based approach to approximate the asymptotic variance and distribution function of two-stage estimators. We study the case of M-estimations in the second stage and consider a large class of consistent estimators in the first stage. This class includes estimators other than M-estimators as well as nonparametric and Bayesian estimators. Our framework is general as we allow first- and second-stage estimators that are not asymptotically normally distributed. Our approach is easily implementable and suitable for complex models. Unlike resampling methods such as bootstrap, we do not require computing the plug-in estimator multiple times.

• Healthcare Quality by Specialists under a Mixed Compensation System: an Empirical Analysis (whith Damien Echevin and Bernard Fortin)

  Slides

• Quasi-Maximum Likelihood Estimator for Peer Effect Models with Partial Network Data

• Children's Gender and Parent Networks (with Michael Vlassopoulos and Yves Zenou)