An event study plot in STATA - multiple, staggered & temporary treatment with spillover effects

[Dheeya Rizmie]

Hello!

Disclaimer - Apologies for any inconsistency in the posting format, I have read instructions/FAQ but this is still my first post so fear I may have violated any standards!

I am looking to produce an event study plot to check for parallel trends in a generalised DD analysis. The project looks at how a congestion policy impacts to traffic. The policy is introduced in certain zones and is only operational for certain hours of the day and certain days of the week. For example, district A is only affected on Saturdays & Sundays from 9am-5pm while district C is affected on Saturdays for 24 hours. District B (neighboring to district A) has spillover effects from the policy, even though they are not included as part of the congestion policy. In a sense, district B is also being "treated" by the policy but to a different intensity. This naturally leads me to ask if an event study would suit this setup to show parallel trends, as the "control" group is also being treated?

If an event study does suit, I have been struggling with using esplot and event_plot. I believe both commands are designed to plot event studies. I believe both require a balanced panel, which (in theory) I do not have. I have a dataset that is unique by road sensor ID and date-time (DDMMYYY hh:mm:ss), see below for example. Even though I have imputed the hours/days for which we do not have traffic measurements, I cannot force these values to be 0 and they remain missing (we cannot say for sure that there were no cars on road ABC). Due to these missing values, event_plot nor esplot seem to not recognize the dataset to be balanced. Is there a possible solution to produce an event study plot?

*attempt with esplot*
esplot debit, event(PolicyDay)
esplot debit, event(PolicyDay) compare(PolicyZone)

*attempt with event_plot*
ppmlhdfe debit l(0/7).PolicyDay f(0/7).PolicyDay   /// poisson for count variable
    i.Annual_CFDay c.mean_rain c.mean_avetemp, /// controls for car-free day, and weather
    absorb(i.date_dow i.date_MM i.date_YYYY i.holiday i.iu_ac) /// time-fixed effects and road sensor FE
    irr vce(cluster i.iu_ac#date_dow) // clustering two-ways (road sensor and day of week)
    
event_plot,  default_look stub_lag(L#event) stub_lead(F#event) together plottype(scatter)

Fine Details: I am using Stata MP 15. Hourly analysis has 72 million observations & daily analysis is at 3.4 million observations. Example of the hourly dataset below. The daily dataset is the below dataset collapsed by date_DMY. iu_ac is the unique identifier for each road sensor. debit is the number of cars.

* Example generated by -dataex-. To install: ssc install dataex
clear
input int(iu_ac date_DMY) double(datetime debit) byte(PolicyZone PolicyDay    PolicyHour)    float(PolicyZone_Buffer5    PolicyDayB3    PolicyHourB3)    byte    Annual_CFDay    float(mean_rain    mean_temperature)
1642 21003  1.814742e+12    735 0 0 0 1 0 0 0         0    24.325
468 21047 1.8184788e+12    118 0 0 0 1 0 0 0      .225     26.35
5679 21377  1.847016e+12      . 0 0 0 0 0 0 0         0    25.675
4523 20385 1.7613216e+12    650 0 0 0 1 0 0 0         0    15.325
6792 20315 1.7552916e+12   1042 0 0 0 1 0 0 0        .1    21.025
6974 20577 1.7778744e+12    210 0 0 0 0 0 0 0         0      16.5
781 20818 1.7986968e+12     21 0 0 0 0 0 0 0 .01111111      .475
4865 20756 1.7933508e+12      . 0 0 0 1 0 0 0 .07777778    13.975
5482 20445 1.7665236e+12      . 0 0 0 0 0 0 0 .01111111    13.325
1039 20672 1.7861256e+12      . 0 0 0 0 0 0 0         0     24.85
1586 20067  1.733868e+12      . 0 0 0 1 0 0 0  .4222222      10.3
6581 19971  1.725552e+12      . 0 0 0 0 0 0 0         0     23.65
6801 20480 1.7694756e+12      . 0 0 0 0 0 0 0 1.6666666        12
4224 21431  1.851696e+12    699 0 0 0 1 0 0 0 .01111111     19.75
6418 19941 1.7229636e+12  507.6 0 0 0 1 0 0 0  22.41111     23.25
794 20576   1.77777e+12     46 0 0 0 1 0 0 0        .4    18.725
5808 20583 1.7784108e+12    401 0 0 0 1 0 0 0  4.366667    21.375
850 19915 1.7207388e+12      . 0 0 0 0 0 0 0     2.625     19.85
6277 19983 1.7265456e+12     72 0 0 0 1 0 0 0      .025     26.74
6654 20531  1.773918e+12      . 0 0 0 1 0 0 0         0    10.575
6014 21367 1.8461376e+12    681 0 0 0 1 0 0 0 .04444445        31
5320 21262 1.8370692e+12    829 0 0 0 0 0 0 0 .23333333 1.1750001
1905 21387 1.8479052e+12    714 0 0 0 1 0 0 0         0      27.9
65 21046 1.8183816e+12      . 0 0 0 1 1 1 0      .275      26.7
6848 21025  1.816578e+12      . 0 0 0 0 0 0 0 .22222222        22
73 21433 1.8518112e+12   1244 0 0 0 1 0 0 0 .22222222      21.4
5050 20036 1.7311392e+12      . 0 0 0 1 1 1 0 .05555556     13.34
4319 20398 1.7624268e+12    308 0 0 0 1 0 0 0 1.1555556    18.275
192 20540 1.7747316e+12    538 0 0 0 1 0 0 0 1.7555555      12.8
941 20303 1.7541936e+12      . 0 0 0 0 0 0 0 1.2714286    32.925
1041 20924  1.807902e+12    307 0 0 0 0 0 0 0  .8111112     15.05
1149 20214 1.7465184e+12      . 0 0 0 0 0 0 0 1.1777778    16.775
6025 19792 1.7100756e+12      . 0 0 0 1 0 0 0         0  20.03333
6641 20328 1.7563608e+12     60 0 0 0 1 0 0 0         0    23.525
867 21039 1.8177804e+12     71 0 0 0 0 0 0 0       4.5    19.075
4599 20780 1.7954604e+12   1160 0 0 0 1 0 0 0     6.425    11.775
5501 20540 1.7746776e+12      . 0 0 0 0 0 0 0 1.7555555      12.8
4621 20355 1.7587476e+12   2593 0 0 0 1 0 0 0 2.0777779 17.574999
5259 20796 1.7968428e+12   2401 0 0 0 0 0 0 0 .06666667    10.425
5162 19874  1.717182e+12      . 0 0 0 0 0 0 0         0    20.975
4953 20462 1.7679852e+12      . 0 0 0 1 0 0 0 3.6444445    11.175
445 19877 1.7174376e+12 1219.2 0 0 0 1 0 0 0       8.2      20.8
5094 21360 1.8455184e+12     62 0 0 0 1 0 0 0         0    25.075
906 21438 1.8522648e+12    123 0 0 0 1 0 0 0         0    29.625
5315 20788 1.7961552e+12   2325 0 0 0 0 0 0 0         0     6.775
5558 19799 1.7106516e+12      . 0 0 0 0 0 0 0         0     12.15
728 20006 1.7285256e+12     47 0 0 0 0 0 0 0       6.1 17.866667
6924 21374  1.846782e+12      . 0 0 0 1 0 0 0         0     26.85
4301 20448   1.76679e+12      . 0 0 0 0 0 0 0 .02222222    13.475
4730 20867 1.8029556e+12    408 0 0 0 0 0 0 0         0     10.55
6921 20748 1.7926416e+12      . 0 0 0 0 0 0 0 1.5333333    13.525
4537 20638  1.783152e+12      . 0 0 0 1 0 0 0  2.733333     17.85
6623 21217 1.8332136e+12      . 0 0 0 1 0 0 0     .5625     6.475
1479 19900 1.7193744e+12      . 0 0 0 1 0 0 0        .4 25.666666
1430 21534  1.860552e+12     10 0 0 0 0 0 0 0 2.3111112       7.6
5630 21209 1.8325404e+12    527 0 0 0 1 0 0 0       1.6     8.475
424 20465 1.7682192e+12      . 0 0 0 1 0 0 0  .4777778         8
4753 20063   1.73349e+12    205 0 0 0 1 0 0 0         0     8.725
6012 20408 1.7633052e+12      . 0 0 0 0 0 0 0  1.411111     15.35
745 20500 1.7712648e+12   1119 0 0 0 0 0 0 0     .0375      5.85
466 21484  1.856268e+12      . 0 0 0 1 0 0 0         0     9.925
6494 21234 1.8346392e+12      . 0 0 0 0 0 0 0       3.4       5.7
974 20814 1.7984124e+12      . 0 0 0 0 0 0 0 .13333334    11.625
4344 21187 1.8305856e+12    356 0 0 0 1 0 0 0         9    11.425
1575 20276 1.7518464e+12      . 0 0 0 0 0 0 0      .725    26.975
6407 20794 1.7966016e+12    285 0 0 0 1 0 0 0 .02222222      11.2
5320 20055 1.7328348e+12    486 0 0 0 0 0 0 0        .1      15.1
732 19935 1.7224344e+12    380 0 0 0 1 0 0 0         0  27.21667
1447 19943 1.7231544e+12      . 0 0 0 0 0 0 0  5.244444      22.6
5370 20723  1.790532e+12   7710 0 0 0 0 0 0 0         0    20.825
6614 21618 1.8678024e+12      . 0 0 0 1 1 1 0      1.75     13.55
1172 21471 1.8551088e+12      . 0 0 0 0 0 0 0         0    26.075
5370 21024 1.8165024e+12   6113 0 0 0 0 0 0 0  2.788889    20.125
916 21219 1.8333684e+12      . 0 0 0 1 0 0 0         0     3.225
1202 19905 1.7198568e+12    428 0 0 0 0 0 0 0         0     23.42
6957 20770 1.7945316e+12      . 0 0 0 0 0 0 0    3.9375     6.875
599 20196  1.744992e+12      . 0 0 0 1 0 0 0         0     16.75
4573 21620 1.8680328e+12      . 0 0 0 0 0 0 0    2.3375    12.425
5824 20435 1.7656452e+12      . 0 0 0 1 0 0 0 .02222222     7.575
4632 20513 1.7723772e+12      . 0 0 0 0 0 0 0         0       7.5
5310 21594 1.8657612e+12   3614 0 0 0 1 0 0 0         0     13.65
979 21090 1.8222408e+12   1027 0 0 0 0 0 0 0 .23333333      20.1
1193 20785   1.79586e+12    216 0 0 0 0 0 0 0     .0375       9.5
5351 20068 1.7339292e+12      . 0 0 0 0 0 0 0  .3222222     10.48
6022 21066 1.8201024e+12      . 0 0 0 1 0 0 0 .02222222     24.25
6297 19833 1.7136432e+12   1917 0 0 0 0 0 0 0         0    18.525
1761 19880 1.7176644e+12      . 0 0 0 0 0 0 0         0    28.075
5873 21161 1.8283824e+12    665 0 0 0 1 0 0 0 2.2444444      4.75
5222 20940  1.809252e+12    168 0 0 0 0 0 0 0         0    15.975
1326 19822 1.7126604e+12    508 0 0 0 1 0 0 0         0     17.54
1484 21374 1.8467388e+12      . 0 0 0 0 0 0 0         0     26.85
4834 21335 1.8433908e+12      . 0 0 0 1 0 0 0         0    25.775
4932 20445 1.7664912e+12   1498 0 0 0 0 0 0 0 .01111111    13.325
4010 20210 1.7461728e+12    240 0 0 0 1 0 0 0  5.144444    14.275
1156 20778 1.7952516e+12    252 0 0 0 0 0 0 0     7.025    13.375
5567 21384 1.8475956e+12     72 0 0 0 0 0 0 0 .44444445    30.375
6795 21348 1.8445428e+12      . 0 0 0 1 0 0 0 .05555556 18.224998
1638 21255 1.8364536e+12    502 0 0 0 1 0 0 0  1.888889    12.675
1440 20909 1.8065664e+12      . 0 0 0 0 0 0 0     1.875     18.15
6373 21188 1.8307224e+12    224 0 0 0 0 0 0 0  6.244444     13.85
end
format %td date_DMY
format %tc datetime

[Justin Niakamal]

Welcome to Statalist. You'll probably want to generate the lags and leads by hand. I suggest taking a look at example 2 in the help file for event_plot where the author shows how to create lags and leads for use with reghdfe.

[Dheeya Rizmie]

Thank you! I will follow that.