Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Panel difference-in-differences regression with Poisson distribution

    Hi everyone,

    I'm trying to run a DID regression for panel data with the following command:
    xtdidregress (y firmsize) (did), group(treat) time(post)

    where
    y is the dependent variable
    firmsize is the control variable(not control group),
    did is the interaction of treat and post
    treat is a dummy indicating treatment or control group,
    and post is a dummy indicating that the observation is in the post or pre-treatment period.
    I've found that the dependent variable is a count value, and based on its mean and variance, I believe that a Poisson regression model is appropriate. How can I change the above model to include a Poisson regression? Should I simply use the xtpoisson command as follows? Does it have the DID analysis in itself?
    xtpoisson (y firmsize) (did), group(treat) time(post)
    or should I stick to xtdidregress and add some commands to it?

    Thank you in advance.
Working...
X