Publications

La plupart des informations présentées ci-dessous ont été récupérées via RePEc avec l'aimable autorisation de Christian Zimmermann
Exit Polls and Voter Turnout in the 2017 French ElectionsJournal articleAlberto Grillo et Eva Raiber, Revue économique, Volume Pub. anticipées, Issue 7, pp. 14-31, 2024

Lors des élections françaises, les médias belges et suisses interfèrent régulièrement en publiant des sondages et des prédictions avant la fermeture des bureaux de vote. Nous utilisons la précocité et le degré de confiance inhabituels des sondages au second tour de l’élection présidentielle de 2017 pour étudier leurs effets sur la participation électorale. Notre analyse compare les taux de participation à différents horaires, aux premier et second tours, et par rapport aux élections de 2012 et 2022. Les résultats montrent une baisse significative de la participation après la publication des sondages à la sortie des urnes. L’effet s’élève à 1,1 point de pourcentage dans l’analyse en triples differences avec l’élection de 2022 et il est plus fort dans les départements limitrophes de la Belgique. Nous constatons également un léger effet underdog pouvant réduire la marge de victoire jusqu’à 1 point de pourcentage.

Optimal Transport for Counterfactual Estimation: A Method for Causal InferenceBook chapterArthur Charpentier, Emmanuel Flachaire et Ewen Gallic, In: Optimal Transport Statistics for Economics and Related Topics, Nguyen Ngoc Thach, Vladik Kreinovich, Doan Thanh Ha et Nguyen Duc Trung (Eds.), 2024, pp. 45-89, Springer Nature Switzerland, 2024

Many problems ask a question that can be formulated as a causal question: what would have happened if...? For example, would the person have had surgery if he or she had been Black? To address this kind of questions, calculating an average treatment effect (ATE) is often uninformative, because one would like to know how much impact a variable (such as the skin color) has on a specific individual, characterized by certain covariates. Trying to calculate a conditional ATE (CATE) seems more appropriate. In causal inference, the propensity score approach assumes that the treatment is influenced by $$\boldsymbol{x}$$x, a collection of covariates. Here, we will have the dual view: doing an intervention, or changing the treatment (even just hypothetically, in a thought experiment, for example by asking what would have happened if a person had been Black) can have an impact on the values of $$\boldsymbol{x}$$x. We will see here that optimal transport allows us to change certain characteristics that are influenced by the variable whose effect we are trying to quantify. We propose here a mutatis mutandis version of the CATE, which will be done simply in dimension one by saying that the CATE must be computed relative to a level of probability, associated to the proportion of x (a single covariate) in the control population, and by looking for the equivalent quantile in the test population. In higher dimension, it will be necessary to go through transport, and an application will be proposed on the impact of some variables on the probability of having an unnatural birth (the fact that the mother smokes, or that the mother is Black).