Code:
+------------------------------------------------------------+ | name y date event treated intera~n | |------------------------------------------------------------| 1. | ACC .0279861 04/sep/2018 0 0 0 | 2. | ACC .0202032 04/sep/2018 0 1 0 | 3. | ACC .0263037 05/sep/2018 0 0 0 | 4. | ACC .0233199 05/sep/2018 0 1 0 | 5. | ACC .0363558 08/nov/2018 1 0 0 | |------------------------------------------------------------| 6. | ACC .0164535 08/nov/2018 1 1 1 | 7. | ACC .0404181 09/nov/2018 1 0 0 | 8. | ACC .0283537 09/nov/2018 1 1 1 | 9. | BBBB .0254419 04/sep/2018 0 0 0 | 10. | BBBB .0248689 04/sep/2018 0 1 0 | |------------------------------------------------------------| 11. | BBBB .0373516 05/sep/2018 0 0 0 | 12. | BBBB .0107468 05/sep/2018 0 1 0 | 13. | BBBB .0358811 08/nov/2018 1 0 0 | 14. | BBBB .0117112 08/nov/2018 1 1 1 | 15. | BBBB .0575707 09/nov/2018 1 0 0 | |------------------------------------------------------------| 16. | BBBB .0116921 09/nov/2018 1 1 1 | +------------------------------------------------------------+
where:
- name is the name of a person
- event indicates pre-period or post period phase
- treatment indicates whether we are in the morning (=0) or in the afternoon (=1)
- interaction is the product between the two
- y is the y-variable
I want to understand if the policy has an impact or not.
I want to do a difference in difference analysis but I don't want to use xtreg because I can "follow a person" over time and I want to use a 1-to-1 matching.
In other words, I want to match observation 1 with observation 2, 3 with 4 and so on...
Is there any way I can do this? I tried with psmatch2 but without success.
Thank you!
0 Response to Exact matching command with panel data (diff-in-diff analysis)
Post a Comment