Hello!
I'm doing a PSM using -psmatch2-. Instead of matching treatment and control firms once only in the sample period, I'm matching treatment and control firms once every year. Currently, I use the same set of variables to match treatment and control firms, and the code below can help me to achieve this. I wonder how this code can be revised so that I could use different sets of variables for different years to perform PSM? The sample period is 2010 to 2019. Thanks a lot for any suggestions.
My code now for year by year PSM using the same matching variables each year
I'm doing a PSM using -psmatch2-. Instead of matching treatment and control firms once only in the sample period, I'm matching treatment and control firms once every year. Currently, I use the same set of variables to match treatment and control firms, and the code below can help me to achieve this. I wonder how this code can be revised so that I could use different sets of variables for different years to perform PSM? The sample period is 2010 to 2019. Thanks a lot for any suggestions.
My code now for year by year PSM using the same matching variables each year
Code:
forvalues i = 2010(1)2019{
preserve
global XX employee asset age litigation ROA
cap keep if year == `i'
cap set seed 0001
cap gen tmp = runiform()
cap sort tmp
psmatch2 y $XX i.industry i.year, out(performance) neighbor(2) caliper(0.05) common ties ate
pstest $XX, both graph
cap drop if _weight == .
cap save `i'.dta, replace
restore
}
use 2010.dta, clear
forvalues i = 2011(1)2019{
cap append using `i'.dta
}
