Announcement

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

  • Difference in difference (DD) with 3 groups

    Hello,

    I'd like to run an DD analyses with a treatment variable (Yi) that has three groups, specifically 3 states with different policy conditions. Group 1 with no treatment (control group with no policy) while Group 2 (passed weak policy) and Group 3 (passed comprehensive policy) have different treatment levels. I'd like to look at the policy impact on tobacco use in all 3 states. I am familiar with difference in difference with one control group and one treatment group using Stata, but I have never tried anything with 3 groups (or even heard of it). Can anyone point me to resources that can guide me in the next steps, specifically using Stata or help me understand how I might specify the regression framework for this type of modeling? Can't seem to find any helpful thing Stata related online?

  • #2
    No different. Dummy up the different treatments.

    basic DID is

    y = treat post c.treat#c.post

    where you have,

    y = weak comp post c.weak$c.post c.comp#c.post

    Comment


    • #3
      This is great. Thank you!

      Comment

      Working...
      X