model with endogenous dichotomous predictor of continuous dependent variable

[Juan Carlos Rodriguez-Raga]

Which Stata (v15.1) command should I use to estimate the following linear model:

y = b0 + b1z + b2x

where:

y is continuous
z is a dummy which is very likely endogenous
x is a vector of control variables

I would like to instrumentalize z with some other predictors. Since z is binary, I should not use ivregress 2sls, right? Then, which command should I use?

I'd really appreciate any help. Thanks!

[FernandoRios]

2sls is perfectly fine
however if you want to model the endogenous variable as a probit, you could use etregress
this command assumes an endogenous treatment effect, which is a binary variable.
hth