Hi all,
I am using ivregress to conduct a instrumental variable estimation (2sls).
I am using the ivregress command in Stata, and have set up the regression as follows:
In the above regression, I have two endogeneous variables (r2011 (which takes the value 0 or 1) and r2011 interacted with an indicator variable for 'guar_boom_dist' (whether a village is located in a boom district). I have respectively instrumented these variables with the instrument t and t interacted with the boom district village.
However, when I run this regression, the ivregress command does not seem to recognise my choice of base order. I specified the bases using the ib0. command, but Stata is instead taking the level 1 for the guar_boom_district indicator as the base. Is there any way I can force Stata to use my base categories?
I am using ivregress to conduct a instrumental variable estimation (2sls).
I am using the ivregress command in Stata, and have set up the regression as follows:
Code:
foreach var in cultiv manlab {
ivregress 2sls nco2d_`var'_share (ib0.r2011 ib0.r2011#ib0.guar_boom_dist = ib0.t ib0.t#ib0.guar_boom_dist ) left right $my_controls i.dist_id [aw = kernel_${kernel}_${mainband}] if mainsample, robust
}
However, when I run this regression, the ivregress command does not seem to recognise my choice of base order. I specified the bases using the ib0. command, but Stata is instead taking the level 1 for the guar_boom_district indicator as the base. Is there any way I can force Stata to use my base categories?
