Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • IV for interaction term

    Dear All,

    I have a question about the implementation of 2SLS regression when the endogenous variable interacts with another exogenous variable.

    For example, assume Y = X + B + X*B where X is endogenous, B is exogenous, and Z is an instrument for X.

    In this case, the following command should be run in Stata:

    Code:
      ivreghdfe Y B (X X*B  =  Z Z*B), absorb(F1 F2)
    However, if X is subsumed by the fixed effects, should the Z be included together with Z*B as an instrument for X*B? Or the Z*B should be the only instrument?
Working...
X