Hi,
I have been trying to estimate different models of a binary outcome variable (non-claim). I used different models as per the esttab command below once treating the main variable of interest (the amount of benefit) as exogenous and once instrumenting it. In the 2sls regression, I had to use the partial out due to the singleton dummy variable error. I show below only the results of the instrumented variable to keep it simple here.
Looking at the significance of the instrumented variable across the different models, it is significant (p<0.05) when the IV probit is unweighted but turns insignificant when it is weighted (model 3). I found both instrumental variables are highly significant in the first stage. Is there anything that I should pay attention to here? I'd better stick to using weights to correct for underrepresentation of population groups in this sample. I just don't know how to deal with different results for the instrumented variable when using the sample weights. Thanks in advance!
I have been trying to estimate different models of a binary outcome variable (non-claim). I used different models as per the esttab command below once treating the main variable of interest (the amount of benefit) as exogenous and once instrumenting it. In the 2sls regression, I had to use the partial out due to the singleton dummy variable error. I show below only the results of the instrumented variable to keep it simple here.
Code:
/*weighted iv (model 3)*/ ivprobit nclaim age age2 married ib3.edu i.eu_nat i.pgemplst_gen i.howner i.singlep i.dis_dummy i.female i.east ib2.citysize i.haskids (benamt = ptransfers needs) [pw=hweight], first
Code:
ivreg2 nclaim age age2 married ib3.edu i.eu_nat i.pgemplst_gen i.howner i.singlep i.dis_dummy i.female i.east ib2.citysize i.haskids (ben_amt = ptransfers needs) [pw=hweight] if head==1, partial(married i.howner i.singlep i.dis_dummy i.east i.haskids
Code:
esttab probit melogit ivprobit ivprobitunweight iv2sls using myregfile.doc, star stats(r2 N)
Code:
--------------------------------------------------------------------------------------------
(1) (2) (3) (4) (5)
--------------------------------------------------------------------------------------------
main
ben_amt -0.0168*** -0.0179*** -0.0317 -0.0210* -0.00786
(-3.82) (-3.71) (-1.74) (-2.42) (-1.58)
--------------------------------------------------------------------------------------------
r2 0.243
N 4239 4239 4239 4239 4239
--------------------------------------------------------------------------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001
