Dear all, I have daily data across months that includes dates in which election is conducted. The date of election varies from state to state. I intend to find out whether holding of election has impacted upon the public service delivery in districts of election state compared to non-election state. There are four states in which election was conducted during this period. I have defined the treatment variable as Posti =1 for the ith election state for all the dates succeding the election date. I then estimated a TWFE model
public service=a+bPost+time fixed effects +district fixed effects, cluster (vce cluster state).
I have the following questions 1) Did I code the treatment variable (Post) correctly and
2) Given the recent litretaure documenting the biases TWFE model what would be the appropriate estimation method.
Finally 3) How to check common trend assumption in TWFE model ?
public service=a+bPost+time fixed effects +district fixed effects, cluster (vce cluster state).
I have the following questions 1) Did I code the treatment variable (Post) correctly and
2) Given the recent litretaure documenting the biases TWFE model what would be the appropriate estimation method.
Finally 3) How to check common trend assumption in TWFE model ?
Comment