Announcement

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

  • Event study using age cohorts (eventdd)

    Hello everyone,


    I am looking at the impact of a large health program-related political scandal on vaccine status among different age groups. I used the following code:

    Code:
    gen timetotreat = yearofbirth - 2006 // 2006 is the year of the event
    
    eventdd vacc_status  $controls2  , hdfe absorb(i.province) timevar(timetotreat) ci(rcap) clust(province) noend accum leads(9) lags(7) graph_op(ytitle("Vaccine Hesitancy") scheme(s1mono))
    I get the following result:

    Click image for larger version

Name:	vaccine hesit.jpg
Views:	1
Size:	20.9 KB
ID:	1729518

    I infer from this that there is no impact and I am observing a time trend. So I added a time trend:

    Code:
    eventdd vacc_status  $controls2 i.year , hdfe absorb(i.province) timevar(timetotreat) ci(rcap) clust(province) noend accum leads(9) lags(7) graph_op(ytitle("Vaccine Hesitancy") scheme(s1mono))
    However, my output produces no SEs, no p-values, nothing. And the event study graph becomes the following:

    Click image for larger version

Name:	vaccine hesit2.jpg
Views:	1
Size:	18.5 KB
ID:	1729519

    Could somebody tell me if I am doing something wrong here?
Working...
X