Coefplot for event study

[Giulia Ferrero]

Hi all,

I am running the following regression:

reghdfe outcome ib1941.year##c.A , absorb(countyID year) cluster(countyID)
coefplot, baselevels omitted vertical keep(*.year#c.A) levels(95 90) order("-6" "-5" "-4" "-3" "-2" "-1" "0") recast(connected) ciopts(recast() color(black black) lwidth(vthin medthick)) yline(0, lpattern(dash) lcolor(black)) xline(8, lcolor(black) lwidth(thin)) lwidth(thin) lcolor(black) msize(small) mcolor(black) msymbol(O) xlabel(1 "1934" 2 "1935" 3 "1936" 4 "1937" 5 "1938" 6 "1939" 7 "1940" 8 "1941" 9 "1942" 10 "1943" 11 "1944" 12 "1945" 13 "1946" 14 "1947" 15 "1948" 16 "1949", labsize(vsmall)) graphregion(fcolor(white) ifcolor(white) ilcolor(white)) ylabel(, labsize(vsmall) nogrid)

and I get a beautiful event study chart with the coefficients estimates centered around 1941.

Now I want to run:

reghdfe outcome ib1941.year##c.A ib1941.year##c.B , absorb(countyID year) cluster(countyID)

and plot on the same chart the coefficients for the two interactions. I want to get two overlapping time series of estimates, one for the interaction with A and the other for B, to compare the estimates over time.

How can I adjust the coefplot code?

Thank you!

[Andrew Musau]

coefplot is from SSC (FAQ Advice #12). Please provide a data example using dataex that is sufficient to run your regression. Otherwise, create a reproducible example using one of Stata's datasets.