did_multiplegt_old
Table of contents
Notes
This is now a legacy command and should be replaced by the much faster did_multiplegt_dyn.
- Based on: Chaisemartin and D’Haultfœuille 2020, 2021
- Program version (if available): -
- Last checked: Nov 2024
Installation
The following installation now adds a collection of various did_multiplegt commands:
ssc install did_multiplegt, replace
Take a look at the help file:
help did_multiplegt_old
Test the command
Please make sure that you generate the data using the script given here
Let’s try the basic did_multiplegt_old command:
did_multiplegt_old Y id t D, robust_dynamic dynamic(10) placebo(10) breps(20) cluster(id) seed(0)
and we get this output:
DID estimators of the instantaneous treatment effect, of dynamic treatment effects if the dynamic option is used, and of placebo tests of the parallel trends assumption if the placebo option is
used. The estimators are robust to heterogeneous effects, and to dynamic effects if the robust_dynamic option is used.
| Estimate SE LB CI UB CI N Switchers
-------------+-----------------------------------------------------------------
Effect_0 | -.0608394 .2767628 -.6032945 .4816157 78 23
Effect_1 | 8.49767 .3463686 7.818787 9.176552 78 23
Effect_2 | 17.64773 .4428714 16.7797 18.51576 78 23
Effect_3 | 25.9377 .5286248 24.90159 26.9738 78 23
Effect_4 | 34.62362 .7949031 33.06561 36.18163 75 23
Effect_5 | 42.85682 .9838843 40.92841 44.78524 64 19
Effect_6 | 51.93103 1.263135 49.45529 54.40677 64 19
Effect_7 | 60.13327 1.582643 57.03129 63.23525 64 19
Effect_8 | 68.82446 1.71335 65.4663 72.18263 64 19
Effect_9 | 77.30792 1.879409 73.62428 80.99156 64 19
Effect_10 | 85.78878 2.259642 81.35988 90.21768 55 19
Average | 40.79851 .9506058 38.93532 42.6617 762 229
Placebo_1 | .1308918 .5697715 -.9858602 1.247644 78 23
Placebo_2 | -.0635463 .2533317 -.5600764 .4329839 78 23
Placebo_3 | -.1275425 .4031795 -.9177744 .6626893 78 23
Placebo_4 | -.3848304 .371244 -1.112469 .3428078 78 23
Placebo_5 | -.4583827 .2356492 -.9202551 .0034897 75 23
Placebo_6 | -.1875761 .3825025 -.9372809 .5621287 64 19
Placebo_7 | .1194069 .335752 -.5386669 .7774807 64 19
Placebo_8 | .0628537 .3785411 -.6790868 .8047943 64 19
Placebo_9 | -.1943704 .3368038 -.854506 .4657651 64 19
Placebo_10 | -.2936845 .4187626 -1.114459 .5270903 64 19
The command also produces by default an event-study graph, unless the firstdiff_placebo option is specified: in that case, we do not recommend putting together first-difference placebos and long-difference event-study estimates on the same event-study graph.
We can also plot the results using the event_plot (ssc install event_plot, replace) command as follows:
event_plot e(estimates)#e(variances), default_look ///
graph_opt(xtitle("Periods since the event") ytitle("Average causal effect") ///
title("did_multiplegt") xlabel(-10(1)10)) stub_lag(Effect_#) stub_lead(Placebo_#) together
and we get this figure:
