**cross country panel regressions with country fixed effects and year fixed effects, and robust standard errors clustered at the country level to allow for serial correlation over time
*millegis
recode millegis 99=.
recode millegis 77=.
reg ccs_dynamic millegis i.cowcode i.year, robust cluster(cowcode)
*statistically significant at the 99.99+% level.

*emergency
recode emer 99=.
reg ccs_dynamic emer i.cowcode i.year, robust cluster(cowcode)


*emergency but with provisions involving the military in declaring or approving a state of emergency
recode Emerdcap 2=1
recode Emerdcap 99=.
recode Emerdcap 88=.
reg ccs_dynamic Emerdcap i.cowcode i.year, robust cluster(cowcode)

*noamnesty regress the effect of the prohibition of amnesty for the military
recode noamnesty 2=0
recode noamnesty 77=.
recode noamnesty 99=.
reg ccs_dynamic noamnesty i.cowcode i.year, robust cluster(cowcode)
*statistically significant at the 99.99+% level for improving civilian control.

*restart the dataset
use "C:\Users\Apinop\OneDrive - University of Virginia\Final Dissertation Draft\Military_CCS_Merged.dta"

tsset cowcode year

*noamnesty regress the effect of amnesty clauses for the military
recode noamnesty 1=0
recode noamnesty 2=1
recode noamnesty 77=.
recode noamnesty 99=.
reg ccs_dynamic noamnesty i.cowcode i.year, robust cluster(cowcode)
*not statistically significant

*creating a regression table
outreg2 using myreg.doc, replace ctitle(Model 1)