relationship between coefficients on the interactions of treatment and time dummies and coeffcients on treatment*post

[Jiaman Xu]

I have the following data set where I have a number of treat and control firms indicated by the "treat" dummy.

* Example generated by -dataex-. For more info, type help dataex
clear
input float(year month) str6 year_month float(treat ind3_code year_month_c y)
2019  1 "20191"  0  9  6    3.533689
2019 10 "201910" 0  9 15    4.623012
2019 11 "201911" 0  9 16    1.274461
2019 12 "201912" 0  9 17    7.730598
2019  2 "20192"  0  9  7    6.124292
2019  3 "20193"  0  9  8   10.031496
2019  4 "20194"  0  9  9    8.790927
2019  5 "20195"  0  9 10    4.740323
2019  6 "20196"  0  9 11    .4423816
2019  7 "20197"  0  9 12    .4193151
2019  8 "20198"  0  9 13    .0819123
2019  9 "20199"  0  9 14    4.857184
2020  1 "20201"  0  9 18    9.590326
2020 10 "202010" 0  9 27    6.798239
2020 11 "202011" 0  9 28    8.679753
2020 12 "202012" 0  9 29   4.5498023
2020  2 "20202"  0  9 19    9.591087
2020  3 "20203"  0  9 20    6.939636
2020  4 "20204"  0  9 21   .14198092
2020  5 "20205"  0  9 22   1.9564253
2020  6 "20206"  0  9 23    5.456525
2020  7 "20207"  0  9 24    7.099039
2020  8 "20208"  0  9 25   1.1000764
2020  9 "20209"  0  9 26    8.961788
2021  1 "20211"  0  9 30    .8228391
2021  2 "20212"  0  9 31     1.19692
2021  3 "20213"  0  9 32    7.815795
2018 10 "201810" 0 27  3    1.825695
2018 11 "201811" 0 27  4   1.2361627
2018 12 "201812" 0 27  5    6.957359
2018  8 "20188"  0 27  1    2.109008
2018  9 "20189"  0 27  2   2.2285697
2019  1 "20191"  0 27  6      3.3959
2019  2 "20192"  0 27  7    3.302672
2019  3 "20193"  0 27  8   3.9273214
2019  4 "20194"  0 27  9 .0042833816
2019  5 "20195"  0 27 10     7.27686
2019  6 "20196"  0 27 11   1.3466753
2019  7 "20197"  0 27 12    .4504642
2019  8 "20198"  0 27 13    8.953969
2019  9 "20199"  0 27 14    6.018908
2018 10 "201810" 0 27  3    6.866826
2018 11 "201811" 0 27  4    7.163794
2018 12 "201812" 0 27  5    6.703581
2018  8 "20188"  0 27  1   4.5353284
2018  9 "20189"  0 27  2   3.3916595
2019  1 "20191"  0 27  6    5.828989
2019 10 "201910" 0 27 15    8.903518
2019 11 "201911" 0 27 16     6.33075
2019 12 "201912" 0 27 17    .6888227
2019  2 "20192"  0 27  7     7.06424
2019  3 "20193"  0 27  8    8.162328
2019  4 "20194"  0 27  9 -.010218575
2019  5 "20195"  0 27 10   3.5556576
2019  6 "20196"  0 27 11    8.572945
2019  7 "20197"  0 27 12   2.4939535
2019  8 "20198"  0 27 13    2.678189
2019  9 "20199"  0 27 14    7.687457
2020  1 "20201"  0 27 18    6.320674
2020 10 "202010" 0 27 27    5.351401
2020 11 "202011" 0 27 28    5.024953
2020 12 "202012" 0 27 29    6.084361
2020  2 "20202"  0 27 19    8.056747
2020  3 "20203"  0 27 20    8.768842
2020  4 "20204"  0 27 21   2.4430835
2020  5 "20205"  0 27 22    8.275479
2020  6 "20206"  0 27 23    8.087458
2020  7 "20207"  0 27 24    .4250393
2020  8 "20208"  0 27 25    4.748792
2020  9 "20209"  0 27 26   2.1541874
2021  1 "20211"  0 27 30    7.691447
2021  2 "20212"  0 27 31    3.012689
2021  3 "20213"  0 27 32    .6683018
2020  1 "20201"  0  9 18    3.717996
2020 10 "202010" 0  9 27    3.632449
2020 11 "202011" 0  9 28    2.878982
2020 12 "202012" 0  9 29    6.638168
2020  2 "20202"  0  9 19   4.4057684
2020  3 "20203"  0  9 20    9.801064
2020  4 "20204"  0  9 21    3.331873
2020  5 "20205"  0  9 22   1.1952995
2020  6 "20206"  0  9 23     1.51032
2020  7 "20207"  0  9 24    6.938107
2020  8 "20208"  0  9 25   .01384309
2020  9 "20209"  0  9 26     8.90509
2018 10 "201810" 0  9  3   3.7120464
2018 11 "201811" 0  9  4   2.5781066
2018 12 "201812" 0  9  5    7.455093
2018  8 "20188"  0  9  1    9.563288
2018  9 "20189"  0  9  2    7.263426
2019  1 "20191"  0  9  6    5.416348
2019 10 "201910" 0  9 15    .6352155
2019 11 "201911" 0  9 16    7.628274
2019 12 "201912" 0  9 17    1.420605
2019  2 "20192"  0  9  7    2.402516
2019  3 "20193"  0  9  8    8.643309
2019  4 "20194"  0  9  9    7.888723
2019  5 "20195"  0  9 10     7.90661
2019  6 "20196"  0  9 11    9.748446
2019  7 "20197"  0  9 12     6.66818
2019  8 "20198"  0  9 13    6.678521
2019  9 "20199"  0  9 14    9.204957
2020  1 "20201"  0  9 18    3.003242
2020 10 "202010" 0  9 27    9.428153
2020 11 "202011" 0  9 28   1.1929506
2020 12 "202012" 0  9 29   1.9244084
2020  2 "20202"  0  9 19    7.753702
2020  3 "20203"  0  9 20    9.057234
2020  4 "20204"  0  9 21      7.3886
2020  5 "20205"  0  9 22    8.408674
2020  6 "20206"  0  9 23    7.513569
2020  7 "20207"  0  9 24   2.3045537
2020  8 "20208"  0  9 25   1.7342184
2020  9 "20209"  0  9 26    .8647256
2021  1 "20211"  0  9 30    6.329988
2021  2 "20212"  0  9 31   2.3335297
2021  3 "20213"  0  9 32    3.225954
2018 10 "201810" 0  9  3   4.3705716
2018 11 "201811" 0  9  4    .8469349
2018 12 "201812" 0  9  5    1.638725
2018  8 "20188"  0  9  1    9.790879
2018  9 "20189"  0  9  2   3.0363605
2019  1 "20191"  0  9  6    4.441983
2019 10 "201910" 0  9 15   10.102244
2019 11 "201911" 0  9 16     4.31952
2019 12 "201912" 0  9 17    7.665234
2019  2 "20192"  0  9  7   1.0445601
2019  3 "20193"  0  9  8    8.930622
2019  4 "20194"  0  9  9   .19009824
2019  5 "20195"  0  9 10    6.310043
2019  6 "20196"  0  9 11    8.884526
2019  7 "20197"  0  9 12    3.014095
2019  8 "20198"  0  9 13  -.06914096
2019  9 "20199"  0  9 14    8.473163
2020  1 "20201"  0  9 18    2.706152
2020 10 "202010" 0  9 27    .6434497
2020  2 "20202"  0  9 19   .17356275
2020  3 "20203"  0  9 20    8.248734
2020  4 "20204"  0  9 21    7.783752
2020  5 "20205"  0  9 22    5.746327
2020  6 "20206"  0  9 23     5.56095
2020  7 "20207"  0  9 24    5.834881
2020  8 "20208"  0  9 25    4.959742
2020  9 "20209"  0  9 26    6.530029
2018 10 "201810" 1 40  3    7.754977
2018 11 "201811" 1 40  4   .02894541
2018 12 "201812" 1 40  5    8.085847
2018  8 "20188"  1 40  1    6.232292
2018  9 "20189"  1 40  2   2.3495069
2019  1 "20191"  1 40  6    3.721237
2019 10 "201910" 1 40 15    5.076848
2019 11 "201911" 1 40 16     7.38858
2019 12 "201912" 1 40 17    5.625891
2019  2 "20192"  1 40  7    .8325879
2019  3 "20193"  1 40  8    7.172107
2019  4 "20194"  1 40  9    3.282642
2019  5 "20195"  1 40 10     5.44907
2019  6 "20196"  1 40 11    6.815018
2019  7 "20197"  1 40 12    1.898395
2019  8 "20198"  1 40 13    3.766986
2019  9 "20199"  1 40 14   4.0202003
2020  1 "20201"  1 40 18   4.5864644
2020 10 "202010" 1 40 27    5.455174
2020 11 "202011" 1 40 28     8.51354
2020 12 "202012" 1 40 29    .7753707
2020  2 "20202"  1 40 19    5.887803
2020  3 "20203"  1 40 20      6.6939
2020  4 "20204"  1 40 21    6.328744
2020  5 "20205"  1 40 22    8.807063
2020  6 "20206"  1 40 23   4.5260687
2020  7 "20207"  1 40 24   2.0946603
2020  8 "20208"  1 40 25    6.620865
2020  9 "20209"  1 40 26    4.398757
2021  1 "20211"  1 40 30       .9259
2021  2 "20212"  1 40 31    2.935745
2021  3 "20213"  1 40 32    .4075056
2018 10 "201810" 0 25  3    5.976005
2018 11 "201811" 0 25  4   .11457504
2018 12 "201812" 0 25  5    8.798559
2018  8 "20188"  0 25  1    .3006993
2018  9 "20189"  0 25  2     6.46421
2019  1 "20191"  0 25  6   3.6124635
2019 10 "201910" 0 25 15   2.7865186
2019 11 "201911" 0 25 16    4.890226
2019 12 "201912" 0 25 17    2.799066
2019  2 "20192"  0 25  7    7.151778
2019  3 "20193"  0 25  8    8.567761
2019  4 "20194"  0 25  9    8.678613
2019  5 "20195"  0 25 10    2.365629
2019  6 "20196"  0 25 11    7.685627
2019  7 "20197"  0 25 12   3.9277806
2019  8 "20198"  0 25 13   4.6925335
2019  9 "20199"  0 25 14   1.6191373
2020  1 "20201"  0 25 18    8.263336
2020 10 "202010" 0 25 27    .4201417
2020 11 "202011" 0 25 28    2.672808
2020  2 "20202"  0 25 19   8.0656595
2020  3 "20203"  0 25 20   3.7176204
2020  4 "20204"  0 25 21    3.485199
2020  5 "20205"  0 25 22    8.199381
2020  6 "20206"  0 25 23   1.8450886
2020  7 "20207"  0 25 24    1.535846
2020  8 "20208"  0 25 25    1.499506
2020  9 "20209"  0 25 26    7.224507
2018 10 "201810" 0  9  3    3.775818
2018 11 "201811" 0  9  4    5.735529
2018 12 "201812" 0  9  5   -.1520811
2018  8 "20188"  0  9  1    3.010874
2018  9 "20189"  0  9  2    4.193386
2019  1 "20191"  0  9  6    8.902131
2019 10 "201910" 0  9 15    7.275138
2019 11 "201911" 0  9 16    5.849213
2019 12 "201912" 0  9 17    6.337542
2019  2 "20192"  0  9  7   2.1444368
2019  3 "20193"  0  9  8    2.903483
2019  4 "20194"  0  9  9    2.609072
2019  5 "20195"  0  9 10    5.675027
2019  6 "20196"  0  9 11    5.631131
2019  7 "20197"  0  9 12    3.372003
2019  8 "20198"  0  9 13    8.216531
2019  9 "20199"  0  9 14    5.875797
2020  1 "20201"  0  9 18    7.452524
2020 10 "202010" 0  9 27    8.750829
2020 11 "202011" 0  9 28   2.7384744
2020 12 "202012" 0  9 29    7.821986
2020  2 "20202"  0  9 19   1.5615367
2020  3 "20203"  0  9 20   .21987346
2020  4 "20204"  0  9 21    8.504027
2020  5 "20205"  0  9 22     6.46848
2020  6 "20206"  0  9 23   4.2816224
2020  7 "20207"  0  9 24    6.255497
2020  8 "20208"  0  9 25    6.721773
2020  9 "20209"  0  9 26   1.2665868
2021  1 "20211"  0  9 30   4.0677843
2021  2 "20212"  0  9 31     6.15018
2021  3 "20213"  0  9 32    9.127183
2018 10 "201810" 1 40  3    .2397677
2018 11 "201811" 1 40  4    9.098163
2018 12 "201812" 1 40  5    6.878891
2018  8 "20188"  1 40  1    7.705475
2018  9 "20189"  1 40  2    8.059814
2019  1 "20191"  1 40  6    5.669284
2019 10 "201910" 1 40 15   .56963056
2019 11 "201911" 1 40 16    2.466423
2019 12 "201912" 1 40 17   1.6816995
2019  2 "20192"  1 40  7   10.272449
2019  3 "20193"  1 40  8    7.688681
2019  4 "20194"  1 40  9    6.915909
2019  5 "20195"  1 40 10    9.170772
2019  6 "20196"  1 40 11    2.443267
2019  7 "20197"  1 40 12    2.751082
2019  8 "20198"  1 40 13    .4396245
2019  9 "20199"  1 40 14    3.085705
2020  1 "20201"  1 40 18    .7060227
2020 10 "202010" 1 40 27   4.4707074
2020 11 "202011" 1 40 28    2.010524
2020 12 "202012" 1 40 29    9.291284
2020  2 "20202"  1 40 19    .6889345
2020  3 "20203"  1 40 20   .59329623
2020  4 "20204"  1 40 21    5.089862
2020  5 "20205"  1 40 22   1.2854624
2020  6 "20206"  1 40 23    8.669716
2020  7 "20207"  1 40 24    .3937246
2020  8 "20208"  1 40 25    9.525309
2020  9 "20209"  1 40 26     9.81271
2021  1 "20211"  1 40 30   .12010994
2021  2 "20212"  1 40 31    8.718388
2021  3 "20213"  1 40 32   1.1474758
2018 10 "201810" 0  9  3    1.108712
2018 11 "201811" 0  9  4    .9231041
2018 12 "201812" 0  9  5    6.171976
2018  8 "20188"  0  9  1    .9404076
2018  9 "20189"  0  9  2   2.5976355
2019  1 "20191"  0  9  6    9.639495
2019 10 "201910" 0  9 15    4.865177
2019 11 "201911" 0  9 16    6.734227
2019 12 "201912" 0  9 17    2.870469
2019  2 "20192"  0  9  7    7.840401
2019  3 "20193"  0  9  8   4.4260254
2019  4 "20194"  0  9  9    6.653377
2019  5 "20195"  0  9 10    6.535165
2019  6 "20196"  0  9 11    6.822266
2019  7 "20197"  0  9 12    3.978481
2019  8 "20198"  0  9 13    7.675358
2019  9 "20199"  0  9 14  -.03120381
2018 10 "201810" 1 40  3    7.969313
2018 11 "201811" 1 40  4    5.566231
2018 12 "201812" 1 40  5    4.687722
2018  8 "20188"  1 40  1   4.3496456
2018  9 "20189"  1 40  2   2.1796577
2019  1 "20191"  1 40  6     7.61565
2019 10 "201910" 1 40 15    9.683228
2019 11 "201911" 1 40 16    5.749743
2019 12 "201912" 1 40 17   3.6171374
2019  2 "20192"  1 40  7      8.5262
2019  3 "20193"  1 40  8     9.51534
2019  4 "20194"  1 40  9   4.1396112
2019  5 "20195"  1 40 10    5.134404
2019  6 "20196"  1 40 11     5.64292
2019  7 "20197"  1 40 12    5.106074
2019  8 "20198"  1 40 13    7.083689
2019  9 "20199"  1 40 14   .12375187
2020  1 "20201"  1 40 18    9.956807
2020 10 "202010" 1 40 27    5.937571
2020 11 "202011" 1 40 28    3.127404
2020 12 "202012" 1 40 29    6.380744
2020  2 "20202"  1 40 19   3.5631514
2020  3 "20203"  1 40 20    9.622343
2020  4 "20204"  1 40 21   8.9865055
2020  5 "20205"  1 40 22    .8544613
2020  6 "20206"  1 40 23    9.180718
2020  7 "20207"  1 40 24   1.6766287
2020  8 "20208"  1 40 25   2.1812272
2020  9 "20209"  1 40 26      7.1474
2021  1 "20211"  1 40 30     9.03538
2021  2 "20212"  1 40 31    2.497206
2021  3 "20213"  1 40 32   1.9131882
2018 10 "201810" 0  1  3   4.4823704
2018 11 "201811" 0  1  4    9.317528
2018 12 "201812" 0  1  5    7.195023
2018  8 "20188"  0  1  1    3.842241
2018  9 "20189"  0  1  2    .6312331
2018 10 "201810" 1 40  3    9.280899
2018 11 "201811" 1 40  4    7.627875
2018 12 "201812" 1 40  5    5.679721
2018  8 "20188"  1 40  1    8.361551
2018  9 "20189"  1 40  2    4.188936
2019  1 "20191"  1 40  6    .7811348
2019 10 "201910" 1 40 15      5.7349
2019 11 "201911" 1 40 16    6.967733
2019 12 "201912" 1 40 17    1.511556
2019  2 "20192"  1 40  7    5.202242
2019  3 "20193"  1 40  8   1.9068187
2019  4 "20194"  1 40  9    4.615948
2019  5 "20195"  1 40 10    .5793773
2019  6 "20196"  1 40 11    5.363412
2019  7 "20197"  1 40 12  -.07862775
2019  8 "20198"  1 40 13    7.311556
2019  9 "20199"  1 40 14   1.7404075
2020  1 "20201"  1 40 18    4.772023
2020 10 "202010" 1 40 27   3.6372724
2020 11 "202011" 1 40 28   3.3409014
2020 12 "202012" 1 40 29    .8243132
2020  2 "20202"  1 40 19    8.984507
2020  3 "20203"  1 40 20    .2682127
2020  4 "20204"  1 40 21   1.2340344
2020  5 "20205"  1 40 22    .8207581
2020  6 "20206"  1 40 23    7.217927
2020  7 "20207"  1 40 24    1.363565
2020  8 "20208"  1 40 25   2.3041565
2020  9 "20209"  1 40 26   1.7743312
2021  1 "20211"  1 40 30   2.0359025
2021  2 "20212"  1 40 31    9.058302
2021  3 "20213"  1 40 32    6.461065
2018 10 "201810" 0  9  3   1.2212433
2018 11 "201811" 0  9  4    8.242736
2018 12 "201812" 0  9  5    4.951109
2018  8 "20188"  0  9  1     3.32671
2018  9 "20189"  0  9  2   2.1232464
2019  1 "20191"  0  9  6   1.0521011
2019 10 "201910" 0  9 15    7.857421
2019 11 "201911" 0  9 16    9.303766
2019 12 "201912" 0  9 17    .7711078
2019  2 "20192"  0  9  7    4.850563
2019  3 "20193"  0  9  8    7.608701
2019  4 "20194"  0  9  9    7.259295
2019  5 "20195"  0  9 10    8.144084
2019  6 "20196"  0  9 11    8.721997
2019  7 "20197"  0  9 12    2.586372
2019  8 "20198"  0  9 13    6.047177
2019  9 "20199"  0  9 14   4.7197566
2020  1 "20201"  0  9 18   1.5225674
2020 10 "202010" 0  9 27   .23214938
2020 11 "202011" 0  9 28    8.297063
2020 12 "202012" 0  9 29   3.3773944
2020  2 "20202"  0  9 19    8.513958
2020  3 "20203"  0  9 20    .4246383
2020  4 "20204"  0  9 21    7.678414
2020  5 "20205"  0  9 22    2.579587
2020  6 "20206"  0  9 23    9.477227
2020  7 "20207"  0  9 24    9.387574
2020  8 "20208"  0  9 25    7.986716
2020  9 "20209"  0  9 26    1.244926
2021  1 "20211"  0  9 30    5.723133
2018 10 "201810" 0  9  3    2.393986
2018 11 "201811" 0  9  4    9.882178
2018 12 "201812" 0  9  5    2.714946
2018  8 "20188"  0  9  1   1.2831987
2018  9 "20189"  0  9  2    .3200619
2019  1 "20191"  0  9  6    1.667642
2019 10 "201910" 0  9 15     .717705
2019 11 "201911" 0  9 16    7.285248
2019 12 "201912" 0  9 17   3.2716775
2019  2 "20192"  0  9  7    7.044349
2019  3 "20193"  0  9  8   2.8273525
2019  4 "20194"  0  9  9    5.384037
2019  5 "20195"  0  9 10    6.807844
2019  6 "20196"  0  9 11    6.575269
2019  7 "20197"  0  9 12    3.509325
2019  8 "20198"  0  9 13    2.745626
2019  9 "20199"  0  9 14    2.954362
2020  1 "20201"  0  9 18    8.606144
2020 10 "202010" 0  9 27     7.99339
2020 11 "202011" 0  9 28    5.505187
2020 12 "202012" 0  9 29    7.860959
2020  2 "20202"  0  9 19   1.3955384
2020  3 "20203"  0  9 20   .05578054
2020  4 "20204"  0  9 21    .6702116
2020  5 "20205"  0  9 22    4.942846
2020  6 "20206"  0  9 23    4.386099
2020  7 "20207"  0  9 24   1.1759576
2020  8 "20208"  0  9 25    8.343729
2020  9 "20209"  0  9 26    4.780529
2021  1 "20211"  0  9 30   2.1436427
2021  2 "20212"  0  9 31    4.890048
2021  3 "20213"  0  9 32     1.85541
2018 10 "201810" 0  9  3   1.1456699
2018 11 "201811" 0  9  4    1.547843
2018 12 "201812" 0  9  5    5.317666
2018  8 "20188"  0  9  1    7.045587
2018  9 "20189"  0  9  2     3.06757
2019  1 "20191"  0  9  6   2.3010592
2019 10 "201910" 0  9 15   2.3323898
2019 11 "201911" 0  9 16    .9371019
2019 12 "201912" 0  9 17    8.257097
2019  2 "20192"  0  9  7   1.6937568
2019  3 "20193"  0  9  8    8.270033
2019  4 "20194"  0  9  9    2.054113
2019  5 "20195"  0  9 10   2.1731987
2019  6 "20196"  0  9 11    8.765087
2019  7 "20197"  0  9 12    4.474798
2019  8 "20198"  0  9 13    2.765065
2019  9 "20199"  0  9 14    4.147869
2020  1 "20201"  0  9 18    1.706628
2020 10 "202010" 0  9 27   4.5809913
2020 11 "202011" 0  9 28    8.912842
2020 12 "202012" 0  9 29   .07903917
2020  2 "20202"  0  9 19   4.4814944
2020  3 "20203"  0  9 20   1.5171982
2020  4 "20204"  0  9 21    5.082428
2020  5 "20205"  0  9 22    2.932981
2020  6 "20206"  0  9 23    6.061301
2020  7 "20207"  0  9 24    9.238971
2020  8 "20208"  0  9 25    3.631132
2020  9 "20209"  0  9 26    9.930732
2021  1 "20211"  0  9 30    1.793153
2021  2 "20212"  0  9 31    3.658489
2021  3 "20213"  0  9 32    2.836752
2018 10 "201810" 0  1  3     .392072
2018 11 "201811" 0  1  4    6.068316
2018 12 "201812" 0  1  5    9.385101
2018  8 "20188"  0  1  1    .6518202
2018  9 "20189"  0  1  2     7.19049
2019  1 "20191"  0  1  6    .3608057
2019 10 "201910" 0  1 15    6.722629
2019 11 "201911" 0  1 16    5.151308
2019 12 "201912" 0  1 17    5.644791
2019  2 "20192"  0  1  7    8.417883
2019  3 "20193"  0  1  8    9.809309
2019  4 "20194"  0  1  9    2.299094
2019  5 "20195"  0  1 10     5.49826
2019  6 "20196"  0  1 11    9.364619
2019  7 "20197"  0  1 12    1.226144
2019  8 "20198"  0  1 13    5.618271
2019  9 "20199"  0  1 14    .6483236
2020  1 "20201"  0  1 18    .7361569
2020  2 "20202"  0  1 19   1.6322197
2020  3 "20203"  0  1 20    3.463088
2020  4 "20204"  0  1 21    3.822316
2020  5 "20205"  0  1 22    9.216046
2018 10 "201810" 0  1  3   1.3424745
2018 11 "201811" 0  1  4    6.949234
2018 12 "201812" 0  1  5     5.70474
2018  8 "20188"  0  1  1    .4438354
2018  9 "20189"  0  1  2   2.9509246
2019  1 "20191"  0  1  6   2.2191827
2019  2 "20192"  0  1  7    7.999144
2019  3 "20193"  0  1  8    9.051006
2019  4 "20194"  0  1  9    7.874681
2019  5 "20195"  0  1 10    .5263755
2018 10 "201810" 0  9  3    8.046903
2018 11 "201811" 0  9  4    8.277057
2018 12 "201812" 0  9  5    5.795954
2018  8 "20188"  0  9  1    2.537629
2018  9 "20189"  0  9  2   2.0996914
2019  1 "20191"  0  9  6    8.316992
2019 10 "201910" 0  9 15     6.21144
2019 11 "201911" 0  9 16   1.9922885
2019 12 "201912" 0  9 17    1.752224
2019  2 "20192"  0  9  7     .823098
2019  3 "20193"  0  9  8    8.692416
2019  4 "20194"  0  9  9    8.126968
2019  5 "20195"  0  9 10    6.761196
2019  6 "20196"  0  9 11    4.622118
2019  7 "20197"  0  9 12    6.943169
2019  8 "20198"  0  9 13    5.715853
2019  9 "20199"  0  9 14    9.578526
2020  1 "20201"  0  9 18    1.889126
2020 10 "202010" 0  9 27     2.16206
2020 11 "202011" 0  9 28     4.94149
end

I have two shocks and I created two dummy variables to indicate them

* first shock (2019.8-2020.3)
gen post = 0
replace post = 1 if year == 2019 & month >= 8
replace post = 1 if year == 2020 & month <= 3

* second shock (2020.4-2021-3)
gen post2 = 0
replace post2 = 1 if year == 2020 & month > 3

gen treat_post = treat * post
gen treat_post2 = treat * post2

* generate time dummies
tabulate year_month_c, generate(t)
sort year month

* interact time dummies with the treatment variable for months in the post period
foreach var of varlist t13-t32{
gen `var'_treat = `var'*treat
}

Then I run the following two regressions

reghdfe y treat_post treat_post2 , a(ind3_code year_month_c i.ind3_code#c.year_month_c) cluster(ind3_code)
reghdfe y  t13_treat - t32_treat, a(ind3_code year_month_c  i.ind3_code#c.year_month_c) cluster(ind3_code)

ind3_code, year_month_c, and i.ind3_code#c.year_month_c controls for industry, year fixed effects, and industry-specific time trends.

I expect that the coefficient on "treat_post" will be approximately equal to the weighted average of the coefficients on the variables from t13_treat to t20_treat, and the coefficient on treat_post2 will be approximately equal to the weighted average of the coefficients on the variables from t21_treat to t32_treat.

However, the results were not as expected. I'm confused especially about the coefficient on t21_treat to t32_treat. They are mostly positive and statistically significant, whereas the estimated coefficient on treat_post2 is negative. Is my expectation wrong, or is there something wrong with the codes?

[George Ford]

You've got repeated dates for ind3_code.




clear
input float ind3_code str6 year_month
9 "201810"
9 "201810"
9 "201810"
9 "201810"
9 "201810"
9 "201810"
9 "201810"
9 "201810"
9 "201811"
9 "201811"
9 "201811"
9 "201811"
9 "201811"
9 "201811"
9 "201811"
9 "201811"
9 "201812"
9 "201812"
9 "201812"
9 "201812"
9 "201812"
9 "201812"
9 "201812"
9 "201812"
9 "20188"
9 "20188"
9 "20188"
9 "20188"
9 "20188"
9 "20188"
9 "20188"
9 "20188"
9 "20189"
9 "20189"
9 "20189"
9 "20189"
9 "20189"
9 "20189"
9 "20189"
9 "20189"
9 "20191"
9 "20191"
9 "20191"
9 "20191"
9 "20191"
9 "20191"
9 "20191"
9 "20191"
9 "20191"
9 "201910"
9 "201910"
9 "201910"
9 "201910"
9 "201910"
9 "201910"
9 "201910"
9 "201910"
9 "201910"
9 "201911"
9 "201911"
9 "201911"
9 "201911"
9 "201911"
9 "201911"
9 "201911"
9 "201911"
9 "201911"
9 "201912"
9 "201912"
9 "201912"
9 "201912"
9 "201912"
9 "201912"
9 "201912"
9 "201912"
9 "201912"
9 "20192"
9 "20192"
9 "20192"
9 "20192"
9 "20192"
9 "20192"
9 "20192"
9 "20192"
9 "20192"
9 "20193"
9 "20193"
9 "20193"
9 "20193"
9 "20193"
9 "20193"
9 "20193"
9 "20193"
9 "20193"
9 "20194"
9 "20194"
9 "20194"
9 "20194"
9 "20194"
9 "20194"
end
[/CODE]

[Jiaman Xu]

Hi George, thanks for your answer! The original data set is at the company-year level but the shocks are at the industry level. I've added a column of firm IDs (id). But I had the same problem when I ran the following:

reghdfe y treat_post treat_post2 , a(id year_month_c i.id#c.year_month_c) cluster(id)
reghdfe y  t13_treat - t32_treat, a(id year_month_c  i.id#c.year_month_c) cluster(id)

Data with firm ids

* Example generated by -dataex-. For more info, type help dataex
clear
input float(year month) str6 year_month float(treat ind3_code year_month_c y id)
2019  1 "20191"  0  9  6     7.55207  1
2019 10 "201910" 0  9 15    .9786136  1
2019 11 "201911" 0  9 16    1.402468  1
2019 12 "201912" 0  9 17    2.684008  1
2019  2 "20192"  0  9  7    7.617968  1
2019  3 "20193"  0  9  8    9.706656  1
2019  4 "20194"  0  9  9    1.361796  1
2019  5 "20195"  0  9 10    8.378426  1
2019  6 "20196"  0  9 11    7.030532  1
2019  7 "20197"  0  9 12    5.920337  1
2019  8 "20198"  0  9 13    8.963018  1
2019  9 "20199"  0  9 14    5.672304  1
2020  1 "20201"  0  9 18    9.840922  1
2020 10 "202010" 0  9 27    8.086982  1
2020 11 "202011" 0  9 28    8.508053  1
2020 12 "202012" 0  9 29    .7565481  1
2020  2 "20202"  0  9 19    3.660541  1
2020  3 "20203"  0  9 20  .009630573  1
2020  4 "20204"  0  9 21    8.275305  1
2020  5 "20205"  0  9 22    3.318109  1
2020  6 "20206"  0  9 23    9.232048  1
2020  7 "20207"  0  9 24    .7320248  1
2020  8 "20208"  0  9 25   8.2603655  1
2020  9 "20209"  0  9 26    4.463721  1
2021  1 "20211"  0  9 30     7.05222  1
2021  2 "20212"  0  9 31   1.1408604  1
2021  3 "20213"  0  9 32    4.813996  1
2018 10 "201810" 0 27  3    9.004827  2
2018 11 "201811" 0 27  4    8.984754  2
2018 12 "201812" 0 27  5    .7181233  2
2018  8 "20188"  0 27  1    5.532055  2
2018  9 "20189"  0 27  2    4.310487  2
2019  1 "20191"  0 27  6    7.552517  2
2019  2 "20192"  0 27  7    7.244119  2
2019  3 "20193"  0 27  8    4.320543  2
2019  4 "20194"  0 27  9    5.175291  2
2019  5 "20195"  0 27 10      6.6696  2
2019  6 "20196"  0 27 11   3.4046366  2
2019  7 "20197"  0 27 12   3.7914956  2
2019  8 "20198"  0 27 13     .335326  2
2019  9 "20199"  0 27 14    6.844749  2
2018 10 "201810" 0 27  3    1.844738  3
2018 11 "201811" 0 27  4     9.65198  3
2018 12 "201812" 0 27  5    9.988294  3
2018  8 "20188"  0 27  1    7.598196  3
2018  9 "20189"  0 27  2    9.392284  3
2019  1 "20191"  0 27  6    1.938892  3
2019 10 "201910" 0 27 15    8.883771  3
2019 11 "201911" 0 27 16   1.8023335  3
2019 12 "201912" 0 27 17    6.246184  3
2019  2 "20192"  0 27  7    3.241234  3
2019  3 "20193"  0 27  8    9.145295  3
2019  4 "20194"  0 27  9    7.471214  3
2019  5 "20195"  0 27 10    5.354929  3
2019  6 "20196"  0 27 11     2.42676  3
2019  7 "20197"  0 27 12    5.298104  3
2019  8 "20198"  0 27 13   2.3757844  3
2019  9 "20199"  0 27 14    9.496355  3
2020  1 "20201"  0 27 18    5.968487  3
2020 10 "202010" 0 27 27   2.1778405  3
2020 11 "202011" 0 27 28    9.300996  3
2020 12 "202012" 0 27 29   3.3564036  3
2020  2 "20202"  0 27 19    2.814975  3
2020  3 "20203"  0 27 20    2.903293  3
2020  4 "20204"  0 27 21    4.842035  3
2020  5 "20205"  0 27 22   4.0251045  3
2020  6 "20206"  0 27 23    8.111124  3
2020  7 "20207"  0 27 24    1.374862  3
2020  8 "20208"  0 27 25    7.263938  3
2020  9 "20209"  0 27 26    8.219784  3
2021  1 "20211"  0 27 30    4.933696  3
2021  2 "20212"  0 27 31    6.391519  3
2021  3 "20213"  0 27 32     7.70928  3
2020  1 "20201"  0  9 18   1.6516296  4
2020 10 "202010" 0  9 27    3.131789  4
2020 11 "202011" 0  9 28    2.896417  4
2020 12 "202012" 0  9 29    7.532889  4
2020  2 "20202"  0  9 19    8.182489  4
2020  3 "20203"  0  9 20     7.68029  4
2020  4 "20204"  0  9 21    2.950384  4
2020  5 "20205"  0  9 22    7.499077  4
2020  6 "20206"  0  9 23   1.0585207  4
2020  7 "20207"  0  9 24    8.960072  4
2020  8 "20208"  0  9 25    5.990088  4
2020  9 "20209"  0  9 26    7.187502  4
2018 10 "201810" 0  9  3    5.763564  5
2018 11 "201811" 0  9  4    3.144333  5
2018 12 "201812" 0  9  5   4.1464753  5
2018  8 "20188"  0  9  1   1.6662228  5
2018  9 "20189"  0  9  2    7.416208  5
2019  1 "20191"  0  9  6    .7370343  5
2019 10 "201910" 0  9 15    4.953378  5
2019 11 "201911" 0  9 16    8.671056  5
2019 12 "201912" 0  9 17    8.441866  5
2019  2 "20192"  0  9  7    6.772014  5
2019  3 "20193"  0  9  8      2.9125  5
2019  4 "20194"  0  9  9   2.1364968  5
2019  5 "20195"  0  9 10    3.219365  5
2019  6 "20196"  0  9 11    5.236912  5
2019  7 "20197"  0  9 12    3.310858  5
2019  8 "20198"  0  9 13    9.474156  5
2019  9 "20199"  0  9 14    .1019091  5
2020  1 "20201"  0  9 18     9.01666  5
2020 10 "202010" 0  9 27    5.089784  5
2020 11 "202011" 0  9 28   3.4462876  5
2020 12 "202012" 0  9 29    .6595117  5
2020  2 "20202"  0  9 19    8.809928  5
2020  3 "20203"  0  9 20    2.636956  5
2020  4 "20204"  0  9 21    6.032826  5
2020  5 "20205"  0  9 22    9.485242  5
2020  6 "20206"  0  9 23     7.61779  5
2020  7 "20207"  0  9 24    9.101017  5
2020  8 "20208"  0  9 25    .3709091  5
2020  9 "20209"  0  9 26   2.0348547  5
2021  1 "20211"  0  9 30   2.3807757  5
2021  2 "20212"  0  9 31    2.590331  5
2021  3 "20213"  0  9 32    7.111765  5
2018 10 "201810" 0  9  3    5.752823  6
2018 11 "201811" 0  9  4     5.80062  6
2018 12 "201812" 0  9  5      6.2007  6
2018  8 "20188"  0  9  1     5.21272  6
2018  9 "20189"  0  9  2   2.9357924  6
2019  1 "20191"  0  9  6     1.32492  6
2019 10 "201910" 0  9 15    7.636866  6
2019 11 "201911" 0  9 16   2.6273155  6
2019 12 "201912" 0  9 17    .6142684  6
2019  2 "20192"  0  9  7   2.9587605  6
2019  3 "20193"  0  9  8     .803114  6
2019  4 "20194"  0  9  9   4.4250474  6
2019  5 "20195"  0  9 10    5.894524  6
2019  6 "20196"  0  9 11    1.402627  6
2019  7 "20197"  0  9 12    8.952939  6
2019  8 "20198"  0  9 13    8.509423  6
2019  9 "20199"  0  9 14    5.345918  6
2020  1 "20201"  0  9 18    3.247553  6
2020 10 "202010" 0  9 27    8.725818  6
2020  2 "20202"  0  9 19    1.586611  6
2020  3 "20203"  0  9 20    .7001554  6
2020  4 "20204"  0  9 21   2.1862614  6
2020  5 "20205"  0  9 22     1.72393  6
2020  6 "20206"  0  9 23    4.814452  6
2020  7 "20207"  0  9 24    9.822144  6
2020  8 "20208"  0  9 25   3.6019285  6
2020  9 "20209"  0  9 26    5.898798  6
2018 10 "201810" 1 40  3   2.4476566  7
2018 11 "201811" 1 40  4     7.90943  7
2018 12 "201812" 1 40  5    1.333635  7
2018  8 "20188"  1 40  1   3.4857955  7
2018  9 "20189"  1 40  2    9.140632  7
2019  1 "20191"  1 40  6    9.478208  7
2019 10 "201910" 1 40 15    3.893855  7
2019 11 "201911" 1 40 16    7.672158  7
2019 12 "201912" 1 40 17     8.77275  7
2019  2 "20192"  1 40  7   2.6259296  7
2019  3 "20193"  1 40  8    9.531048  7
2019  4 "20194"  1 40  9    1.588546  7
2019  5 "20195"  1 40 10    6.790689  7
2019  6 "20196"  1 40 11    7.126815  7
2019  7 "20197"  1 40 12    2.458712  7
2019  8 "20198"  1 40 13   1.6764313  7
2019  9 "20199"  1 40 14    5.254681  7
2020  1 "20201"  1 40 18    7.016855  7
2020 10 "202010" 1 40 27   2.1484842  7
2020 11 "202011" 1 40 28   2.1745336  7
2020 12 "202012" 1 40 29    4.408948  7
2020  2 "20202"  1 40 19   1.4903368  7
2020  3 "20203"  1 40 20    6.821652  7
2020  4 "20204"  1 40 21   2.0739052  7
2020  5 "20205"  1 40 22    4.953641  7
2020  6 "20206"  1 40 23    .2307996  7
2020  7 "20207"  1 40 24    2.977416  7
2020  8 "20208"  1 40 25    5.949654  7
2020  9 "20209"  1 40 26    3.944947  7
2021  1 "20211"  1 40 30    6.177876  7
2021  2 "20212"  1 40 31   4.6472473  7
2021  3 "20213"  1 40 32    5.234149  7
2018 10 "201810" 0 25  3    8.329607  8
2018 11 "201811" 0 25  4    3.899847  8
2018 12 "201812" 0 25  5    9.370722  8
2018  8 "20188"  0 25  1    9.354629  8
2018  9 "20189"  0 25  2    3.017852  8
2019  1 "20191"  0 25  6    8.485166  8
2019 10 "201910" 0 25 15    5.164359  8
2019 11 "201911" 0 25 16    7.660375  8
2019 12 "201912" 0 25 17   .06122523  8
2019  2 "20192"  0 25  7    9.302339  8
2019  3 "20193"  0 25  8   3.7246194  8
2019  4 "20194"  0 25  9    .4260934  8
2019  5 "20195"  0 25 10   2.1328244  8
2019  6 "20196"  0 25 11    9.883374  8
2019  7 "20197"  0 25 12    8.072899  8
2019  8 "20198"  0 25 13    5.263667  8
2019  9 "20199"  0 25 14    1.667525  8
2020  1 "20201"  0 25 18    1.230856  8
2020 10 "202010" 0 25 27    9.841234  8
2020 11 "202011" 0 25 28    2.145945  8
2020  2 "20202"  0 25 19   4.6190553  8
2020  3 "20203"  0 25 20   1.4640977  8
2020  4 "20204"  0 25 21    5.176486  8
2020  5 "20205"  0 25 22    5.716005  8
2020  6 "20206"  0 25 23    8.191779  8
2020  7 "20207"  0 25 24    5.814484  8
2020  8 "20208"  0 25 25    4.054936  8
2020  9 "20209"  0 25 26    6.696227  8
2018 10 "201810" 0  9  3    4.838048  9
2018 11 "201811" 0  9  4   2.0370777  9
2018 12 "201812" 0  9  5   1.4334027  9
2018  8 "20188"  0  9  1    9.695601  9
2018  9 "20189"  0  9  2    7.663661  9
2019  1 "20191"  0  9  6    1.653828  9
2019 10 "201910" 0  9 15    7.076856  9
2019 11 "201911" 0  9 16     1.19275  9
2019 12 "201912" 0  9 17    9.374759  9
2019  2 "20192"  0  9  7    6.847514  9
2019  3 "20193"  0  9  8    7.272093  9
2019  4 "20194"  0  9  9    9.466451  9
2019  5 "20195"  0  9 10    3.423412  9
2019  6 "20196"  0  9 11    9.084236  9
2019  7 "20197"  0  9 12    1.903814  9
2019  8 "20198"  0  9 13    6.374664  9
2019  9 "20199"  0  9 14    6.005998  9
2020  1 "20201"  0  9 18    6.326343  9
2020 10 "202010" 0  9 27    3.984069  9
2020 11 "202011" 0  9 28    3.718146  9
2020 12 "202012" 0  9 29   3.7028346  9
2020  2 "20202"  0  9 19    3.986612  9
2020  3 "20203"  0  9 20    5.590534  9
2020  4 "20204"  0  9 21    5.770776  9
2020  5 "20205"  0  9 22    7.937202  9
2020  6 "20206"  0  9 23    6.875431  9
2020  7 "20207"  0  9 24    4.880559  9
2020  8 "20208"  0  9 25   1.0597852  9
2020  9 "20209"  0  9 26     4.76765  9
2021  1 "20211"  0  9 30    2.818333  9
2021  2 "20212"  0  9 31    6.244217  9
2021  3 "20213"  0  9 32    1.474202  9
2018 10 "201810" 1 40  3   1.8761444 10
2018 11 "201811" 1 40  4    9.124126 10
2018 12 "201812" 1 40  5    7.891165 10
2018  8 "20188"  1 40  1    7.420677 10
2018  9 "20189"  1 40  2    9.274618 10
2019  1 "20191"  1 40  6    1.951128 10
2019 10 "201910" 1 40 15    8.287621 10
2019 11 "201911" 1 40 16    8.444612 10
2019 12 "201912" 1 40 17    7.288074 10
2019  2 "20192"  1 40  7    8.471776 10
2019  3 "20193"  1 40  8    6.797028 10
2019  4 "20194"  1 40  9   2.8162744 10
2019  5 "20195"  1 40 10    7.250293 10
2019  6 "20196"  1 40 11    5.429574 10
2019  7 "20197"  1 40 12    8.038015 10
2019  8 "20198"  1 40 13     4.77181 10
2019  9 "20199"  1 40 14    .7238613 10
2020  1 "20201"  1 40 18    9.080024 10
2020 10 "202010" 1 40 27    8.216396 10
2020 11 "202011" 1 40 28   1.2984966 10
2020 12 "202012" 1 40 29    7.165498 10
2020  2 "20202"  1 40 19   1.4105202 10
2020  3 "20203"  1 40 20   3.5067585 10
2020  4 "20204"  1 40 21     2.53459 10
2020  5 "20205"  1 40 22     5.98816 10
2020  6 "20206"  1 40 23   2.1863124 10
2020  7 "20207"  1 40 24   2.9850426 10
2020  8 "20208"  1 40 25    8.613802 10
2020  9 "20209"  1 40 26    8.974747 10
2021  1 "20211"  1 40 30   4.1742578 10
2021  2 "20212"  1 40 31    3.761248 10
2021  3 "20213"  1 40 32      .67303 10
2018 10 "201810" 0  9  3    5.859678 11
2018 11 "201811" 0  9  4    4.961307 11
2018 12 "201812" 0  9  5    .1814808 11
2018  8 "20188"  0  9  1    8.835435 11
2018  9 "20189"  0  9  2   1.2851697 11
2019  1 "20191"  0  9  6    4.824552 11
2019 10 "201910" 0  9 15    8.475065 11
2019 11 "201911" 0  9 16    5.527618 11
2019 12 "201912" 0  9 17    6.710842 11
2019  2 "20192"  0  9  7     2.71026 11
2019  3 "20193"  0  9  8     3.72781 11
2019  4 "20194"  0  9  9    4.904513 11
2019  5 "20195"  0  9 10    8.252763 11
2019  6 "20196"  0  9 11    8.414296 11
2019  7 "20197"  0  9 12    6.412801 11
2019  8 "20198"  0  9 13    8.180822 11
2019  9 "20199"  0  9 14    .7455062 11
2018 10 "201810" 1 40  3    5.937301 12
2018 11 "201811" 1 40  4   3.4150515 12
2018 12 "201812" 1 40  5   2.3472867 12
2018  8 "20188"  1 40  1   2.3324742 12
2018  9 "20189"  1 40  2     6.59022 12
2019  1 "20191"  1 40  6    1.375673 12
2019 10 "201910" 1 40 15    3.866729 12
2019 11 "201911" 1 40 16    5.482247 12
2019 12 "201912" 1 40 17    6.148804 12
2019  2 "20192"  1 40  7    3.259921 12
2019  3 "20193"  1 40  8     1.98041 12
2019  4 "20194"  1 40  9    3.212341 12
2019  5 "20195"  1 40 10    .9074852 12
2019  6 "20196"  1 40 11    5.740378 12
2019  7 "20197"  1 40 12    7.293422 12
2019  8 "20198"  1 40 13   .08754096 12
2019  9 "20199"  1 40 14    7.502211 12
2020  1 "20201"  1 40 18    7.160038 12
2020 10 "202010" 1 40 27    8.560918 12
2020 11 "202011" 1 40 28   4.4149003 12
2020 12 "202012" 1 40 29    3.793423 12
2020  2 "20202"  1 40 19    6.899609 12
2020  3 "20203"  1 40 20   .12535164 12
2020  4 "20204"  1 40 21    3.556548 12
2020  5 "20205"  1 40 22    .7354485 12
2020  6 "20206"  1 40 23    6.138276 12
2020  7 "20207"  1 40 24    9.744926 12
2020  8 "20208"  1 40 25     1.35516 12
2020  9 "20209"  1 40 26   1.4342084 12
2021  1 "20211"  1 40 30 .0003548431 12
2021  2 "20212"  1 40 31   1.1851813 12
2021  3 "20213"  1 40 32    4.925122 12
2018 10 "201810" 0  1  3    8.163175 13
2018 11 "201811" 0  1  4     7.39435 13
2018 12 "201812" 0  1  5    4.116225 13
2018  8 "20188"  0  1  1    6.186018 13
2018  9 "20189"  0  1  2    .6203403 13
2018 10 "201810" 1 40  3    8.165086 14
2018 11 "201811" 1 40  4   4.4563217 14
2018 12 "201812" 1 40  5    3.655759 14
2018  8 "20188"  1 40  1     8.88096 14
2018  9 "20189"  1 40  2   10.000744 14
2019  1 "20191"  1 40  6   1.9739002 14
2019 10 "201910" 1 40 15    3.683996 14
2019 11 "201911" 1 40 16   1.2756222 14
2019 12 "201912" 1 40 17     6.41056 14
2019  2 "20192"  1 40  7    4.590661 14
2019  3 "20193"  1 40  8   10.084952 14
2019  4 "20194"  1 40  9    1.563199 14
2019  5 "20195"  1 40 10    8.275483 14
2019  6 "20196"  1 40 11    1.388385 14
2019  7 "20197"  1 40 12   2.3186655 14
2019  8 "20198"  1 40 13    8.641288 14
2019  9 "20199"  1 40 14    .9660628 14
2020  1 "20201"  1 40 18     8.20918 14
2020 10 "202010" 1 40 27    6.739127 14
2020 11 "202011" 1 40 28   10.300123 14
2020 12 "202012" 1 40 29      8.2985 14
2020  2 "20202"  1 40 19   3.1088755 14
2020  3 "20203"  1 40 20    7.193382 14
2020  4 "20204"  1 40 21    8.936744 14
2020  5 "20205"  1 40 22    5.845172 14
2020  6 "20206"  1 40 23   3.1618774 14
2020  7 "20207"  1 40 24   4.6240706 14
2020  8 "20208"  1 40 25     2.52147 14
2020  9 "20209"  1 40 26    9.555015 14
2021  1 "20211"  1 40 30    7.612161 14
2021  2 "20212"  1 40 31    6.726914 14
2021  3 "20213"  1 40 32    6.500872 14
2018 10 "201810" 0  9  3   2.3815832 15
2018 11 "201811" 0  9  4   2.0368853 15
2018 12 "201812" 0  9  5   .12240066 15
2018  8 "20188"  0  9  1    8.829661 15
2018  9 "20189"  0  9  2   1.4159802 15
2019  1 "20191"  0  9  6    1.362707 15
2019 10 "201910" 0  9 15    6.133142 15
2019 11 "201911" 0  9 16   8.6155615 15
2019 12 "201912" 0  9 17    2.831199 15
2019  2 "20192"  0  9  7   4.6036143 15
2019  3 "20193"  0  9  8   3.3924725 15
2019  4 "20194"  0  9  9    .5149167 15
2019  5 "20195"  0  9 10     4.37327 15
2019  6 "20196"  0  9 11    6.132393 15
2019  7 "20197"  0  9 12   3.1126904 15
2019  8 "20198"  0  9 13   2.8333704 15
2019  9 "20199"  0  9 14    1.217176 15
2020  1 "20201"  0  9 18     6.31824 15
2020 10 "202010" 0  9 27    6.232001 15
2020 11 "202011" 0  9 28    9.476906 15
2020 12 "202012" 0  9 29    9.441483 15
2020  2 "20202"  0  9 19    9.505275 15
2020  3 "20203"  0  9 20     6.79731 15
2020  4 "20204"  0  9 21   1.5790466 15
2020  5 "20205"  0  9 22     9.27956 15
2020  6 "20206"  0  9 23    8.533494 15
2020  7 "20207"  0  9 24     6.83346 15
2020  8 "20208"  0  9 25   2.9076855 15
2020  9 "20209"  0  9 26   3.9247525 15
2021  1 "20211"  0  9 30    7.703138 15
2018 10 "201810" 0  9  3    9.065013 16
2018 11 "201811" 0  9  4    3.171321 16
2018 12 "201812" 0  9  5   1.9748538 16
2018  8 "20188"  0  9  1   3.7073326 16
2018  9 "20189"  0  9  2   2.2483292 16
2019  1 "20191"  0  9  6   4.3794627 16
2019 10 "201910" 0  9 15    2.896263 16
2019 11 "201911" 0  9 16   1.1547415 16
2019 12 "201912" 0  9 17   2.1316648 16
2019  2 "20192"  0  9  7    1.761142 16
2019  3 "20193"  0  9  8    7.313239 16
2019  4 "20194"  0  9  9    2.977915 16
2019  5 "20195"  0  9 10    5.930423 16
2019  6 "20196"  0  9 11    9.326112 16
2019  7 "20197"  0  9 12    8.137642 16
2019  8 "20198"  0  9 13    8.925205 16
2019  9 "20199"  0  9 14   1.4414163 16
2020  1 "20201"  0  9 18    3.719126 16
2020 10 "202010" 0  9 27     7.73646 16
2020 11 "202011" 0  9 28     4.09581 16
2020 12 "202012" 0  9 29    6.047992 16
2020  2 "20202"  0  9 19    9.394392 16
2020  3 "20203"  0  9 20    7.183547 16
2020  4 "20204"  0  9 21   4.0604196 16
2020  5 "20205"  0  9 22   3.9876864 16
2020  6 "20206"  0  9 23     .031765 16
2020  7 "20207"  0  9 24    9.062113 16
2020  8 "20208"  0  9 25    6.979237 16
2020  9 "20209"  0  9 26   3.7387085 16
2021  1 "20211"  0  9 30    5.063859 16
2021  2 "20212"  0  9 31    6.030518 16
2021  3 "20213"  0  9 32    .7819452 16
2018 10 "201810" 0  9  3     2.04127 17
2018 11 "201811" 0  9  4    8.617164 17
2018 12 "201812" 0  9  5    3.062447 17
2018  8 "20188"  0  9  1    7.177741 17
2018  9 "20189"  0  9  2    8.005289 17
2019  1 "20191"  0  9  6   2.7469206 17
2019 10 "201910" 0  9 15   4.2824583 17
2019 11 "201911" 0  9 16    6.691942 17
2019 12 "201912" 0  9 17    7.400893 17
2019  2 "20192"  0  9  7   2.6358685 17
2019  3 "20193"  0  9  8   4.0461287 17
2019  4 "20194"  0  9  9    8.169775 17
2019  5 "20195"  0  9 10    5.613908 17
2019  6 "20196"  0  9 11    5.340521 17
2019  7 "20197"  0  9 12    7.046791 17
2019  8 "20198"  0  9 13    4.366727 17
2019  9 "20199"  0  9 14    9.483538 17
2020  1 "20201"  0  9 18   .54515773 17
2020 10 "202010" 0  9 27    3.545572 17
2020 11 "202011" 0  9 28    6.890961 17
2020 12 "202012" 0  9 29    3.288022 17
2020  2 "20202"  0  9 19    8.896478 17
2020  3 "20203"  0  9 20    8.710363 17
2020  4 "20204"  0  9 21    6.760664 17
2020  5 "20205"  0  9 22    .9503378 17
2020  6 "20206"  0  9 23    3.917658 17
2020  7 "20207"  0  9 24    .3847771 17
2020  8 "20208"  0  9 25    4.808871 17
2020  9 "20209"  0  9 26    8.225313 17
2021  1 "20211"  0  9 30    1.518843 17
2021  2 "20212"  0  9 31    3.944212 17
2021  3 "20213"  0  9 32    2.955542 17
2018 10 "201810" 0  1  3    2.496096 18
2018 11 "201811" 0  1  4    6.492787 18
2018 12 "201812" 0  1  5    2.845662 18
2018  8 "20188"  0  1  1   .19175343 18
2018  9 "20189"  0  1  2   3.0073586 18
2019  1 "20191"  0  1  6    9.549002 18
2019 10 "201910" 0  1 15    6.863246 18
2019 11 "201911" 0  1 16    7.690255 18
2019 12 "201912" 0  1 17    3.735141 18
2019  2 "20192"  0  1  7    8.168134 18
2019  3 "20193"  0  1  8   1.2520688 18
2019  4 "20194"  0  1  9    9.686743 18
2019  5 "20195"  0  1 10    7.859705 18
2019  6 "20196"  0  1 11    .5949351 18
2019  7 "20197"  0  1 12    7.013388 18
2019  8 "20198"  0  1 13   1.1217903 18
2019  9 "20199"  0  1 14    4.905401 18
2020  1 "20201"  0  1 18    7.073771 18
2020  2 "20202"  0  1 19    9.595904 18
2020  3 "20203"  0  1 20    6.536124 18
2020  4 "20204"  0  1 21    7.226632 18
2020  5 "20205"  0  1 22    .3510576 18
2018 10 "201810" 0  1  3    4.654211 19
2018 11 "201811" 0  1  4   4.6256423 19
2018 12 "201812" 0  1  5    8.653395 19
2018  8 "20188"  0  1  1    3.642115 19
2018  9 "20189"  0  1  2    8.923897 19
2019  1 "20191"  0  1  6    7.591852 19
2019  2 "20192"  0  1  7    6.418539 19
2019  3 "20193"  0  1  8    1.686367 19
2019  4 "20194"  0  1  9    2.860294 19
2019  5 "20195"  0  1 10    5.917571 19
2018 10 "201810" 0  9  3    5.402507 20
2018 11 "201811" 0  9  4    5.671078 20
2018 12 "201812" 0  9  5   4.5595655 20
2018  8 "20188"  0  9  1    7.704699 20
2018  9 "20189"  0  9  2    .8245859 20
2019  1 "20191"  0  9  6    5.945876 20
2019 10 "201910" 0  9 15   1.7850147 20
2019 11 "201911" 0  9 16    2.887656 20
2019 12 "201912" 0  9 17   2.3033497 20
2019  2 "20192"  0  9  7    6.264809 20
2019  3 "20193"  0  9  8    2.724624 20
2019  4 "20194"  0  9  9     6.62279 20
2019  5 "20195"  0  9 10    6.323635 20
2019  6 "20196"  0  9 11    6.049952 20
2019  7 "20197"  0  9 12  -.13420914 20
2019  8 "20198"  0  9 13    2.954927 20
2019  9 "20199"  0  9 14   2.4882104 20
2020  1 "20201"  0  9 18    6.458018 20
2020 10 "202010" 0  9 27    5.261304 20
2020 11 "202011" 0  9 28    5.965877 20
end

[George Ford]

Interesting problem. I'm reviewing a paper right now that does something similar, so I'm curious as to the answer.

It appears it has something to do with both the absorbed slopes and the imbalanced sample.

I cleaned up your date code.

gen date = ym(year, month)
format %tm date
xtset id date

* first shock (2019.8-2020.3)
g post = tin(2019m8, 2020m3)
* second shock (2020.4-2021-3)
g post2 = tin(2020m4,2021m3)

gen treat_post = treat * post
gen treat_post2 = treat * post2
sort year month

* generate time dummies
tabulate year_month_c, generate(t)
sort year month

* interact time dummies with the treatment variable for months in the post period
foreach var of varlist t13-t32{
    gen `var'_treat = `var'*treat
}

egen Nid = count(id), by(id)
tab Nid

** Balanced, No Slopes
reghdfe y c.treat#c.post c.treat#c.post2 if Nid==32, a(ind3_code date) cluster(ind3_code)
qui reghdfe y t13_treat  - t32_treat if Nid==32, a(ind3_code date) cluster(ind3_code)
lincom (t13 + t14 + t15 + t16 + t17 + t18 + t19 + t20)/8

** Balanced, Add Slopes
reghdfe y c.treat#c.post c.treat#c.post2 if Nid==32, a(ind3_code date i.ind3_code#c.date) cluster(ind3_code)
reghdfe y t13_treat  - t32_treat if Nid==32, a(ind3_code date i.ind3_code#c.date) cluster(ind3_code)
lincom (t13 + t14 + t15 + t16 + t17 + t18 + t19 + t20)/8

** Allow imbalanced, no slopes
reghdfe y c.treat#c.post c.treat#c.post2, a(ind3_code date) cluster(ind3_code)
qui reghdfe y t13_treat  - t32_treat, a(ind3_code date) cluster(ind3_code)
tabstat t13-t20 if e(sample) , save stats(mean sum) format(%8.4f)
lincom (t13*0.034 + t14*0.034 + t15*0.032 + t16*0.032 + t17*0.032 + t18*0.032 + t19*0.03 + t20*0.03)/0.256

** A little unbalanced
reghdfe y c.treat#c.post c.treat#c.post2 if Nid>=30, a(ind3_code date) cluster(ind3_code)
qui reghdfe y t13_treat  - t32_treat if Nid>=30, a(ind3_code date) cluster(ind3_code)
tabstat t13-t20 if e(sample) , save stats(mean sum) format(%8.4f)
lincom (t13 + t14 + t15 + t16 + t17 + t18 + t19 + t20)/8

** More imbalanced
reghdfe y c.treat#c.post c.treat#c.post2 if Nid>=27, a(ind3_code date) cluster(ind3_code)
qui reghdfe y t13_treat  - t32_treat if Nid>=27, a(ind3_code date) cluster(ind3_code)
tabstat t13-t20 if e(sample) , save stats(mean sum) format(%8.4f)
lincom (t13 + t14 + t15 + t16 + t17 + t18 + t19 + t20)/8

[George Ford]

Same with id date

reghdfe y c.treat#c.post c.treat#c.post2 if Nid==32, a(id date) 
qui reghdfe y t13_treat - t32_treat if Nid==32, a(id date)
lincom (t13 + t14 + t15 + t16 + t17 + t18 + t19 + t20)/8

reghdfe y c.treat#c.post c.treat#c.post2 if Nid==32, a(id date i.id#c.date) 
reghdfe y t13_treat - t32_treat if Nid==32, a(id date i.id#c.date)
lincom (t13 + t14 + t15 + t16 + t17 + t18 + t19 + t20)/8

[Clyde Schechter]

I expect that the coefficient on "treat_post" will be approximately equal to the weighted average of the coefficients on the variables from t13_treat to t20_treat, and the coefficient on treat_post2 will be approximately equal to the weighted average of the coefficients on the variables from t21_treat to t32_treat.

No, this expectation is incorrect. It is simply untrue that, in general, the coefficient of x in a regression will be some weighted average of the coefficients of x in regressions conditioned on the value of some other variable.

Here is a simple (simplistic) example showing this:

clear*

set obs 100
gen x = _n
gen group = x <= 50
set seed 1234
gen y = 0.1*x + rnormal()
replace y = 0.01*x + rnormal() + 25 if group

regress y x
regress y c.x##i.group
margins group, dydx(x)

graph twoway lfit y x || lfit y x if group || lfit y x if !group || scatter y x

The graph makes it clear exactly what is going on here. In your data the situation is probably not so stark. But the principle remains: the regression coefficient of x in the full data does not even have to fall within the range of the coefficients conditional on values of other variables, or even have the same sign.

[Jiaman Xu]

No, this expectation is incorrect. It is simply untrue that, in general, the coefficient of x in a regression will be some weighted average of the coefficients of x in regressions conditioned on the value of some other variable.

Hi Clyde, thank you for the reply and the example! I understand that it is not true that the coefficient of x in a regression will be some weighted average of the coefficients of x in regressions conditioned on the value of some other variable, and your example clearly illustrates the point.

But I think my question is different. I was trying the compare the coefficients on Treat*Post and the coefficients on Treat*T1, Treat*T2,...., Treat*Tn. T1, T2,...,Tn are dummy variables indicating the relevant period, and "Post" is set to 1 for periods from T1 to Tn (i.e. Post = T1+T2+T3...+Tn).

I wonder if in this case, the coefficient on Treat*Post will be the weighted average of the coefficients on Treat*T1, Treat*T2,...,Treat*Tn.

Thank you.

[Clyde Schechter]

No, it's not actually a different thing. Consider just the regression on the data during the Post period. Then Treat*Post is just the coefficient of Treat during the Post period minus the coefficient of Treat during the Pre period, and Treat*Ti is the coefficient of Treat in year Ti minus the coefficient in the Pre period. But the post period consists of precisely the Ti years. So this is just like the example I showed with Ti representing group and Treat representing X.

Added: I do note, however, that for this kind of situation to arise there must be a correlation between X and group. If X and group are independent, this could not happen. So, if you had a data set where the panels are all balanced, and where all the panels underwent treatment at the same time, then Treat would be independent of year, and your expectation would be correct. But your data is not like that.

[Jiaman Xu]

No, it's not actually a different thing. Consider just the regression on the data during the Post period. Then Treat*Post is just the coefficient of Treat during the Post period minus the coefficient of Treat during the Pre period, and Treat*Ti is the coefficient of Treat in year Ti minus the coefficient in the Pre period. But the post period consists of precisely the Ti years. So this is just like the example I showed with Ti representing group and Treat representing X.

Added: I do note, however, that for this kind of situation to arise there must be a correlation between X and group. If X and group are independent, this could not happen. So, if you had a data set where the panels are all balanced, and where all the panels underwent treatment at the same time, then Treat would be independent of year, and your expectation would be correct. But your data is not like that.

hi Clyde, thank you for the further explanation. I see now why they are the same thing.

Interesting problem. I'm reviewing a paper right now that does something similar, so I'm curious as to the answer.

It appears it has something to do with both the absorbed slopes and the imbalanced sample.

I cleaned up your date code.

Thank you George for the code and the examples, they are very helpful.


To summarize the replies so far, the problem arises because of the imbalanced sample I have.
It also has something to do with the absorbed linear time trends (i.ind3_code#c.date), because it arises even in a balanced sample (and all panels underwent treatment at the same time) when the slopes are absorbed. But the reason is not clear.

[George Ford]

Thanks Clyde for jumping in.