In the study of the employment effect, Xi
includes age, age squared, education, marital status,
health, household size and other provincial-level macroeconomic indicators. In the analysis
of the working hour effect, Xi
includes age, age squared, wage, education and other provincial
indicators.
As stated above, the core assumption of the DID approach is that the “same outcome
changes without treatment.” In practice, it is difficult to test this hypothesis directly. In the
next section, we further examine the comparability of the experiment and the control groups.