Announcement

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

  • Mundlak in combination with instrumental variable method

    Hello, I am currently working on a project where I am estimating my endogenous independent variable with 2 instruments. The problem is, there instruments do not change over time. I found out that I need to use the fe model when using the Hausman test on my xtreg. However, when I am using the fe with my instrumental variable regression (xtivreg), it is not possible as the instruments do not change over time. I was thinking of using Mundlak, as there you can have variables that do not change over time. But, I do not fully understand how this would work.

    After installing the mundlak package, I figured out you can do mundlak dependent independent and get a sort of xtreg. However, I am not able to figure out how to do this mundlak command in combination with my instrumental variable regression. Can someone help me with this?

    Also, once this works, do I have to do this for my xtreg too or can I just use fe there? As my instruments themselves will not be in this regression?

    Furthermore, once this instrumental variable regression works, how do I see if it is a good one? Most internet sites say to use estat endogenous but this does not work for my stata 17 and I do not seem to understand the manual for postestimation after xtivreg.

    Another random question: if I am doing my xtreg, how do I figure out if I should use vce(robust) or cluster, and if I use cluster, how does this work? Cluster(id)?

    Thank you in advance for the help. I am a stata beginner
Working...
X