Hi all,
I would like to estimate the following equation:
Y = b0 + b1T + X*beta + U1 (1)
where T is a binary variable and Y is a discrete variable (1 to 4); and U1 is an error term. The variable T is endogenous and I have two discrete instrumental variables Z1 and Z2. So, the first stage equation is:
T = Z*gamma + X*theta + U2 (2)
I wonder if I can use the control function approach to estimate equations (1) and (2)? Or does someone have another approach that I can use to have consistent estimates?
Thanks in advance.
Related Posts with Multinomial logit/probit models with binary endogenous explanatory variable
Conditional scatterplot (partial regression plot) for a quadratic relationship- In which parentheses (gmmstyle() or ivstyle()) should I insert my explanatory variables?
- How to count the number of variables in multiple locals and meanwhile to compute the mean value of each group using for loop in Stata?
- Graph: text not centered within textbox
- Seemingly Unrelated Regression with endogenous variables
Subscribe to:
Post Comments (Atom)
0 Response to Multinomial logit/probit models with binary endogenous explanatory variable
Post a Comment