Announcement

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

  • Psmatch2 matched by firm

    Hi all!

    I have a question about psmatch2. Now I am trying to use psmatch2 to generate untreated group for DID analysis, my final sample contains firm-year observations. Now I can perfectly generate 1-1 firm-year matching, for example, firmA in 2001 is matched with firmB in 2001, firmC in 2005 is matched with firmD in 2005. However, the problem is, sometimes, firmC in 2005 is also matched with firmB in 2005, so the numbers of firm in untreated group sometimes will be smaller than the numbers of firm in treated group (since I think my code is to match each firm-year with each firm-year observation). What I want to do is: if firmB is matched to firmA in 2001, then I do not want firmB to be matched with firmC in other year, say 2005, therefore the numbers of firm in treated group and untreated group will be the same.
    Iwould like to askif there are options for psmatch2 to solve this problem? or i should use other methods?

    Many thanks in advance!!!


    Example of my code:

    probit Indicator Ln_Asset Leverage Profitability Tangibility ROA i.fyear, iterate(10)
    predict pscore1 if e(sample), pr
    gen pscore2=fyear*100+pscore1
    psmatch2 Indicator, pscore(pscore2) noreplacement caliper(0.5) comm neighbor(1)


  • #2
    1

    Comment

    Working...
    X