In this section, we will lay out the identification strategy applied in the present paper.
Following the difference-in-differences (DID) approach in Athey and Imbens (2006), all
observations are divided into two groups: an experiment group and a control group. The