2017-09 , Hitotsubashi Institute for Advanced Study, Hitotsubashi University
May 2017; reviced in September 2017 A growing body of literature examines the relationships between historical events and contemporary economic outcomes. Recent studies estimate the causal effects using detailed historical data and contemporary microdata of individuals and/or households. In this paper, we discuss conceptual and econometric issues inherent in the causal inference following the potential outcomes framework. Using an empirical example, we also discuss a simple alternative approach to avoid these issues that is coherent with the potential outcomes framework.