Hi all,
This question is a cross posting from Stack Exchange, with a bit more detail. I have a repeated cross section data on households. I'm trying a DiD by comparing adjacent districts which lie on two sides of state border, before and after a policy. The design I'm aiming for is the following: state A introduced a policy in year t, while state B (adjacent state) did not. Now, state A has a district p (say) which shares a border with districts r and s of state B. The idea is that we compare outcomes between p and r,s before and after period t. So district p after t is treatment group, while r,s form control group. It might also happen that two treatment districts in state A share border with same control district in state B.
While I have come across such research design earlier, I can't think of any specific paper now so that I may look it up.
Here is a sample:
here A is treatment state, B is control state. post: dummy for years after policy.
Now from the map, we know that p shares border with r,s; q shares border with r.
I'm having trouble with preparing the data for this comparison. My DiD would compare p with r and s, and q with r, before and after the policy (given by the variable post), but I can't seem to figure out how to make that happen.
Any help would be appreciated. Please reply if you feel the question needs more detail and I'll add them.
Thanks!
This question is a cross posting from Stack Exchange, with a bit more detail. I have a repeated cross section data on households. I'm trying a DiD by comparing adjacent districts which lie on two sides of state border, before and after a policy. The design I'm aiming for is the following: state A introduced a policy in year t, while state B (adjacent state) did not. Now, state A has a district p (say) which shares a border with districts r and s of state B. The idea is that we compare outcomes between p and r,s before and after period t. So district p after t is treatment group, while r,s form control group. It might also happen that two treatment districts in state A share border with same control district in state B.
While I have come across such research design earlier, I can't think of any specific paper now so that I may look it up.
Here is a sample:
Code:
* Example generated by -dataex-. For more info, type help dataex clear input str3 ID long(state district) byte(x y post) "111" 1 1 1 1 1 "112" 1 1 0 1 1 "113" 1 1 1 0 0 "114" 1 2 0 1 1 "115" 1 2 0 0 0 "116" 1 2 1 0 1 "117" 2 3 0 0 0 "118" 2 3 0 1 1 "119" 2 3 0 0 1 "120" 2 4 0 1 1 "121" 2 4 1 0 1 "122" 2 4 1 1 0 end label values state state1 label def state1 1 "A", modify label def state1 2 "B", modify label values district district1 label def district1 1 "p", modify label def district1 2 "q", modify label def district1 3 "r", modify label def district1 4 "s", modify
Now from the map, we know that p shares border with r,s; q shares border with r.
I'm having trouble with preparing the data for this comparison. My DiD would compare p with r and s, and q with r, before and after the policy (given by the variable post), but I can't seem to figure out how to make that happen.
Any help would be appreciated. Please reply if you feel the question needs more detail and I'll add them.
Thanks!
Comment