Hi everyone,

I am running a difference in differences regression with Panel data on Stata version 14.2.

I have state-level clustered standard errors, but since there are only 32 states in my data I want to use bootstrapping to correct for bias in the standard errors.
Since I am using survey data I need to include weights [pweight]. These are, however, incompatible with the bootstraps option after reg.

I was recommended the boottest command, but I have trouble getting the results out of Stata into a table.

I run the following code:

Code:
 reg violence treat i.state i.year [pweight=factor] if year>2003, cluster(ent)      

matrix v = e(V)    
matrix b = e(b)      

boottest treat=0, boottype(wild) seed(22)    
matrix v[1,1] = (_b[treat]/r(t))^2          

eststo: ereturn post b v    
sum violence if year>2003    
local N    = r(N)    
local mean = r(mean)      
estadd scalar m  = `mean'    
estadd scalar obs     = `N'      

#delimit ;
esttab, b(%-9.3f) se(%-9.3f) noobs keep(treat) nonotes nogaps
mlabels(, none) nonumbers style(tex) fragment replace noline
label stats(obs m, fmt(%9.0gc %5.3f)
label("\\ Observations" "Mean of Dependent Variable"));

#delimit cr
I get the following error message after the line
Code:
   eststo: ereturn post b v
last estimation results not found, nothing to store r(301);
.


Any idea how to fix this? I have been trying around with eststo and naming my estimation but all I get Stata to display is the coefficient&standard error from the OLS regression.


Any help is greatly appreciated.

Mia