Announcement

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

  • using quaids command for two demand equations?

    Hi stata users,
    Quite some time ago I've modelled a two-good demand system (water expenditure // everything else) model using the nlsur quaids command. My analysis focusses entirely on the own-price elasticities that I find for water.
    Now that I've learned about the quaids command that exists in the meantime, I wanted to replicate & update my results, and make use of the much more elaborate post-estimation features.
    However, the quaids command only works for 3 or more goods / budget shares / equations. I can't find the exact reason why (maybe the programmed reporting on cross-elasticities?)
    What would be the adequate way to try to use the quaids functionalities of postestimation for my 2-good demand system?
    Can I insert one good category that I'm not interested in, just in order to have three, or better to adjust the ado files so that it allows for two goods.. I can't really anticipate the problems either route would bring up.
    Thanks for thinking along
    Josefine

  • #2
    Josefine,

    I haven't thought through the economics carefully, but I can confirm that if you make the necessary change to line 65 of quaids.ado, you can fit a two-good demand system. In fact, don't change quaids.ado. Make a copy of that file and name it quaids2.ado. In that new file change line 3 to read
    Code:
    program quaids2, eclass
    and change line 65 to read
    Code:
    if `neqn' < 2 {
    If you're pedantic, you can change the error message in the next line as well. Then call -quaids2- instead of -quaids-.

    The most interesting thing I've found is that the compensated cross-price elasticity of good 1 with respect to the price of good 2 is equal to the negative of the own-price elasticity, and similarly for good 2. I'm sure that just follows from the algebra, though I haven't actually put pencil to paper to verify.
    Last edited by Brian Poi; 11 Jun 2021, 08:22.

    Comment


    • #3
      many thanks, this works indeed and re-assuringly, gives about the same results as my nlsur-try from 10 years back. you are right about the elasticities mirroring each other, the algebra can be simplified considerably with only two goods. Many thanks again, both for the programme and for the fast reply.

      Comment

      Working...
      X