Discussion on "Causal and Counterfactual Views of Missing Data Models"

We sincerely congratulate Profs.Nabi, Bhattacharya, Shpitser, and Robins on their interesting contribution, which addresses challenges in missing data analysis through the framework of potential outcomes.In this paper, the authors explain how missing data problems can be framed as causal inference problems: the complete variables are viewed as counterfactual outcomes, the missingness indicators are treated as treatment variables, and the partially observed variables are interpreted as combinations of potential outcomes and treatments.