eventstudyinteract
Table of contents
Notes
- Based on: Sun and Abraham (2020). Estimating Dynamic Treatment Effects in Event Studies with Heterogeneous Treatment Effects.
- Program version (if available): 0.5
- Last checked: Nov 2024
- Additional info in this blog post.
Installation and options
ssc install eventstudyinteract, replace
Take a look at the help file:
help eventstudyinteract
Test the command
Please make sure that you generate the data using the script given here
Let’s try the basic eventstudyinteract command the never_treated as the control_cohort:
eventstudyinteract Y L_* F_*, vce(cluster id) absorb(id t) cohort(first_treat) control_cohort(never_treat)
which will show this output:
IW estimates for dynamic effects Number of obs = 1,800
Absorbing 2 HDFE groups F(236, 29) = .
Prob > F = .
R-squared = 0.9999
Adj R-squared = 0.9999
Root MSE = 1.0114
(Std. err. adjusted for 30 clusters in id)
------------------------------------------------------------------------------
| Robust
Y | Coefficient std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
L_0 | -.0705566 .4520534 -0.16 0.877 -.9951096 .8539963
L_1 | 8.477542 .4379319 19.36 0.000 7.581871 9.373214
L_2 | 17.69445 .6109211 28.96 0.000 16.44497 18.94392
L_3 | 25.97508 .6262269 41.48 0.000 24.6943 27.25585
L_4 | 34.74591 .9050206 38.39 0.000 32.89493 36.59688
L_5 | 42.5053 1.223893 34.73 0.000 40.00216 45.00845
L_6 | 51.87653 1.498852 34.61 0.000 48.81103 54.94202
L_7 | 60.49124 1.799291 33.62 0.000 56.81128 64.1712
L_8 | 69.1176 1.982205 34.87 0.000 65.06353 73.17167
L_9 | 77.28842 2.229958 34.66 0.000 72.72764 81.8492
L_10 | 85.61498 2.555366 33.50 0.000 80.38867 90.84129
L_11 | 93.99587 2.708476 34.70 0.000 88.45642 99.53533
L_12 | 103.5575 3.047215 33.98 0.000 97.32524 109.7897
<OUTPUT TRUNCATED>
F_44 | .4711243 .7455523 0.63 0.532 -1.053701 1.99595
F_45 | -1.302809 .9750776 -1.34 0.192 -3.297066 .6914491
F_46 | -.1432933 .5311437 -0.27 0.789 -1.229604 .9430176
F_47 | -.5297326 1.077463 -0.49 0.627 -2.733392 1.673927
F_48 | -.3128089 .743813 -0.42 0.677 -1.834077 1.208459
F_49 | -.0590544 .7666954 -0.08 0.939 -1.627123 1.509014
F_50 | -1.156447 .3122062 -3.70 0.001 -1.79498 -.5179134
F_51 | -.2063065 .7599222 -0.27 0.788 -1.760522 1.347909
F_52 | .0657318 .5711765 0.12 0.909 -1.102455 1.233919
F_53 | -.5740659 .5549477 -1.03 0.309 -1.709061 .5609296
F_54 | .3670635 1.075711 0.34 0.735 -1.833013 2.56714
F_55 | -.4353981 .7411865 -0.59 0.561 -1.951295 1.080498
------------------------------------------------------------------------------
In order to plot the estimates we can use the event_plot (ssc install event_plot, replace) command where we restrict the figure to 10 leads and lags:
event_plot e(b_iw)#e(V_iw), default_look graph_opt(xtitle("Periods since the event") ytitle("Average effect") xlabel(-10(1)10) ///
title("eventstudyinteract")) stub_lag(L_#) stub_lead(F_#) trimlag(10) trimlead(10) together
And we get:
