Announcement

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

  • Control variables and fixed effects

    Hello all,

    I have a set of panel data for 15 countries over 10 years and I have included 5 control variables. I have read that control variables are meant to be keep constant and so have created an average value for each variable over the period. However, by doing this, when it comes to running country fixed effects obviously I can’t use these controls because they are all time invariant. Is this normal? Is there something I am missing to change this?

  • #2
    Safia:
    actually I do not remember that control variables should be averaged over to be kept constant in panel data regression with -fe- specification.
    Kind regards,
    Carlo
    (Stata 18.0 SE)

    Comment


    • #3
      Hi Carlo,

      thanks for your reply, are you saying that I do not have to keep them constant or that my method of doing so is wrong? (The other way I saw was to take the value from one year), but again this wouldn’t work with fe.

      Comment


      • #4
        Safia:
        I meant that you do not have to keep them constant (unless they're essentially time-invariant but, if they were so, the -fe- machinery would wipe them, as you stated).
        Kind regards,
        Carlo
        (Stata 18.0 SE)

        Comment

        Working...
        X