This project develops new econometrics and computational tools to make inference on high dimensional models in economics. These models aim at exploiting the informational content present in large data sets with the purpose of addressing economic questions of interest in both microeconomics and macroeconomics. The first contribution of this project is to make prediction and causal inference in economic settings where the effects of counterfactual policies are of interest, like the effects of introducing a new product, of advertisement, or of implementing a government policy. One tool that we propose in this project is the construction of debiased Lasso-type estimators. A second tool is based on the construction of estimators for functional parameters by using information theoretic approaches like Empirical Likelihood and its generalizations. Our second contribution is to exploit sources of information coming from Google search data to nowcast (i.e. “predict the current value of”) quantities of interest for central banks like GDP growth. For this purpose we aim at using factor models where the factors are extracted by using sparse principal component.

(Mise à jour le 10 juillet 2024)

Retour à la liste des projets labellisés par la MSH Paris-Saclay