Announcement

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

  • problem using allsynth by wiltshire and quarterly data

    Hello,

    For my dissertation I am using allsynth to do DiD with unemployment rates. I have both quarterly and yearly data, it is panel data. Using yearly data it works fine but with my quarterly data it has some trouble.
    I use the following code:
    bysort region (qdates): gen date=tq(2005q1)+_n-1
    format date %tq
    xtset regcode date
    drop if regcode ==12
    drop if regcode ==13
    drop if regcode ==1
    drop if regcode ==6
    this is where i encournter problems. I have formatted the data as quarters but i cannot run allsynth like this:
    allsynth UR UR(2014q1) UR(2011q1) UR(2005q1), trperiod(2014q1) trunit(11) figure gapfigure(placebos) pval keep(refugeeresults) plac rep
    This also does not work:
    allsynth UR UR('=tq(2014q1)') UR('=tq(2011q1)') UR('=tq(2005q1)'), trunit(11) trperiod('=tq(2014q1)') figure gapfigure(placebos) pval keep(refugeeresults) plac rep

    The only way it works is if i use the numbers that correspond to quarters:
    allsynth UR UR(216) UR(204) UR(180), trperiod(216) trunit(11) figure gapfigure(placebos) pval keep(refugeeresults) plac rep

    but this means that my graphs are labelled with numbers and not quarters. Any idea how to fix this so the graphs have proper labels?

    Would appreciate any help!



  • #2
    This was an issue in an earlier version of allsynth. Try again with the latest version:
    Code:
    ssc install allsynth, replace all

    Comment


    • #3
      Yep I noticed this (I think) when I was using allsynth, but as Justin says, it's now been fixed.

      Comment

      Working...
      X