0
votes

I am running two regressions in Stata: one without controls and another with controls. I'm using lincom to find the coefficient and se for the sum of two of my regressors. I am then using outreg2 to create an Excel table with my results. So, here is an example of what I'm running:

reg Y var1 var2  var3, cluster(cluster)

local fmodel = Ftail(`e(df_m)',`e(df_r)',`e(F)')
lincom var2+var3
local coef=r(estimate)
local se=r(se)
outreg2 using finlev1.xls , addstat("Joint", `coef', " ", `se',  "F test model", e(F),"P-value of F model", `fmodel') replace ctitle(Title (no controls)) label

reg Y var1 var2  var3 var4 var5, cluster(cluster) 

local fmodel = Ftail(`e(df_m)',`e(df_r)',`e(F)')
lincom var2+var3
local coef=r(estimate)
local se=r(se)
outreg2 using finlev1.xls , addstat("Joint", `coef', " ", `se',  "F test model", e(F),"P-value of F model", `fmodel') append ctitle(Title (with controls)) label

Everything works fine except for one thing: in the resulting table I only get the first regression Standard Error (from lincom), the second one is blank. Actual coefficients from lincom are there, but one of se is missing.

Any ideas why this may be happening?

Thank you!

1

1 Answers

2
votes

outreg2 might be seeing an already existing statistic and thinking that nothing more needs to be done. Try changing the text in addstat.

BTW, lincom var2+var3 is not the the sum of two regressors. It is a sum of the estimated coefficients.