Announcement

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

  • ml command with endogenous variables

    I am estimating utility function parameters with STATAs ml command. I am wondering if there is a way to include an instrumental variable with the ml command. The code (to be run after having defined the ml program) looks like this. Now suppose I have an endogenous variable in the regressors, what can I do? Thank you

    ml model lf RP_CPT (gamma: choice prob1a prob2a prob1b prob2b x1a x2a x1b x2b moment = $det ) (lambda: $det ) (sigma: $det ) (theta: $det ) (noise: ), technique(nr) cluster(id)

  • #2
    Also, if it is not possible, I would be grateful for an answer, thanks.

    Comment


    • #3
      ml is very general, so what is possible or not possible depends on the model, in your case RP_CPT, which we don't have and thus cannot comment on.

      However, there is one general comment we can make: instrumental variables are often estimated with gmm instead of ml. I don't know if it is impossible with ml, but I suspect that it is at the very least easier with gmm.
      ---------------------------------
      Maarten L. Buis
      University of Konstanz
      Department of history and sociology
      box 40
      78457 Konstanz
      Germany
      http://www.maartenbuis.nl
      ---------------------------------

      Comment

      Working...
      X