A0359
Title: Differential error component models for causal identification
Authors: Yongtao Jin - Beijing Institute of Technology (China) [presenting]
Abstract: Fixed and random effects models are popular methods for identifying the causal effects of treatment. First, a class of the generalized least square estimators and their asymptotic properties of the fixed and random effects models are introduced. Then, the differential error component models are proposed that exclude the fixed effects of units and time periods, and the random effects of units and time periods are considered. The treatment effects of heterogeneity analysis and optimal model for a given data are also shown. Meanwhile, the application of a prior study is revisited. The results demonstrate that the estimators of differential random effects models are more significant and robust than the differential, FE, or RE models.
