Hello, I have a treatment at the district level while the outcome is at the individual level, so I have employed district-fixed effects, year-fixed effects, state-by-year fixed effects, and district-specific time trends in my regression. Is this alright?
My understanding is that while district-specific time trends just only capture variation in one direction either it is always up or always down, state-year fixed effects allow us to capture other variation at the state level with the year either up or down. But my question is will the direction that is captured by the district trend be subsumed by the state year fixed effects?
My understanding is that while district-specific time trends just only capture variation in one direction either it is always up or always down, state-year fixed effects allow us to capture other variation at the state level with the year either up or down. But my question is will the direction that is captured by the district trend be subsumed by the state year fixed effects?
Comment