I have been trying to use calloway and Santanna DiD on IPUMS CPS data but not doing so well due to having repeated cross section.

The other day I came across an online thread where it was suggested that if I collapse it on state level then this will be easy to implement due to the nature of repeated cross section. At first I clean and collapse the data at state level in stata then I run the CSDiD on R.

If anyone can kindly give me a feedback on my coding , that's going to make a big difference in further progress of my research.


*** Collapsing on state level ( I collapsed year but I ain't sure that's a right thing to do or not . However, I'll need the year variable for CSDID implementation later )


collapse ln_incwage male ismarried wasmarried age black asian hispanic lths hsdegree somecollege law_change year gvar [aw=asecwt], by(statefip)


Does the collapsing on state level seem reasoanble ?