Causal mediation analysis in economics: Objectives, assumptions, models

Abstract Mediation analysis aims at identifying and evaluating the mechanisms through which treatment affects an outcome. The goal is to disentangle the total treatment effect into two components: the indirect effect that occurs due to one or more intermediate variables, known as mediators, and the direct effect that captures all other possible explanations for why a treatment works. This paper reviews the methodological advancements in the literature on causal mediation in economics, specifically quasi‐experimental designs. I define the parameters of interest, the main assumptions and the identification strategies under the counterfactual approach, and present the Instrumental Variables (IV), Difference‐in‐Differences (DID), and Synthetic Control (SC) methods.