IV with dummy variables

[Xenia Waibel]

Hello everybody,

I am running regressions on Stata. Now I wanted to check whether one of 3 variables could serve as instruments. All of them are dummies. I typed:
ivreg2 var1 (var2 = i.iv1 i.iv2 i.iv3) control1 control 2 control3

Which leads to the following message:
Warning - endogenous variable(s) collinear with instruments
Vars now exogenous: stakorient_score
Warning - collinearities detected
and then it drops all variables.

I played with all kinds of combinations of the IVs, but only one gets me a regression table:
ivreg2 var1 (var2 = i.iv3) control1 control 2 control3
However, even though I get an OLS estimation table, I think it still drops all variables because it says 0 (ommited) in all lines of the table.

Is it not possible to run ivreg2 with dummy instrumentals? Does anyone have a solution to this?

Thanks in advance!

[Phil Bromiley]

You'll increase your chances of a helpful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output, and sample data using dataex. It is a good idea to show us exactly what you entered and what you got. Obviously, the warnings in red show that you did not enter what you have listed in ivreg2.

It is completely possible to have dummy variables as instruments. It looks like var2 is perfectly explained by v1, v2, and v3. If so, then they can't function as instruments.