Difference-in-Difference Matching - Different Intervention for Different HHs at Different Times.

[Sumgait Usagi]

Dear everyone,

I have a survey data. I want to estimate the effect of the survey participation. My question is that in contrast to the classic DID settings of 1 time intervention for the treatment groups, we have a survey and every month different people are signing up (and I also have randomly selected non-participants group for the same period) for an entire year.
Each month new households signup to the program. Program lasted a year. And I would like to evaluate the effect of the program participation. I have monthly 3 years of the data set:
1) year before the Survey,
2) Survey year (for everyone)
3) Post survey year. How can I estimate the ATE for all, instead of particular months?

* Should I calculate ATE for every months separately then take the average? If I do that some months has very few observations --- I am not sure about this.
* Or should I put all the samples (surveyed group), regardless of months, to the same group (Treatment) and calculate ATE - I am not sure about this as well, this would overgeneralize.

I am planning to employ DID Matching with non-participants?
ID= individuals for 36months of data
y=consumption
x=Individual characteristics (age, income, education, zipcode, ethnic group and etc), it also include month and year people signup the surveys.
I assigned survey participants as treatment group. Non-participants are Comparison group.
How can calculate the ATE for the Survey? and How can estimate month after and year after the survey effect?

Please advise. I really need a support!!! I really appreciate for any help.
Kind regards,

[Clyde Schechter]

So, within your treatment group, you already have the ability to distinguish the pre- from the post- intervention period. The absolute point in time differs among the various participants, but for each observation for each person in the treatment group you can identify whether it is pre- or post.

Your difficult arises because you cannot do the same for the controls. There are a couple of ways to proceed.

1. Do you have any information that explains why the people who joined the treatment group at a particular time did so, rather than doing so earlier or later? If you have that information also for the controls, then you can use it to impute a time at which they "would have" entered treatment had they done so, and then calculate pre- and post- from there.

2. If there is no way to do #1, you can just form matched pairs between treatment and control people. Then impute to the matched control the same pre- and post- periods as the treatment person they are matched to. The matching might either be systematic, based on people's attributes, or, if that is not sensible, purely done at random.

Once you have attributed a "would have started treatment" time to each control, and then defined pre- and post- relative to that, then the analysis is just the standard DID approach. You regress the outcome against treatment group, pre-post, and the interaction of those two. The coefficient of the interaction term is the DID estimator of the treatment effect.

This question comes up frequently on this Forum, and if you search on difference-in-differences you will find many examples of precisely this same situation, with detailed explanations and code paradigms.

[Sumgait Usagi]

Dear Clyde,

Thank you for your comments... However, I apologize for not quite understanding your comments. Attached file shows the dates that individuals participated in survey. So, there are multiple "treatment" dates. And there are almost 200 thousand observation of non-participants for the same periods.

So, what you are suggesting that I should create a variable and sign them as treated or not for each survey dates? So, I will have some 40 dummy variables of treatment?
Consumption Dates: 2006m6 - 2014m4. So I have overall post-survey period. Monthly consumption data is at least 36months for each ID.

Using STATA how can I calculate the ATE?
1. Should I create different dummy for each month as "treat" and then estimate ATE for them? And at the end take the mean of all the ATEs. - This doesn't make much sense to me, since I will have very little observation for each "treatment" variable.
2. Or Should I just put everyone who participated the survey and estimate whether overall survey programs affects consumption of goods? - This is also not a feasible approach of putting entire year in the same basket.

Once, I figure out the how to evaluate the ATE issue, I plan to use matching methods as well. Especially, I would like to compare the early participants with later participants, as well as non-participants (Sianesi (2004, 2008), and Kniesner (2016) Working Paper)

I glanced almost all the discussion in the Forum, I could not find one that would help me with the topic and STATA codes as well.
Please help. I really need some guidance.

Thank you very much.

[Clyde Schechter]

Let's back up and take this one step at a time. To do a DID analysis, you need two dichotomous variables (and their interaction).

The first variable is simple: treatment vs control. This variable takes on the same value for every observation of an individual participant: he/she is either in the treatment group (1) or the control group (0) throughout the study.

The second variable is pre_post. In the classic DID design this is also simple. In that design there is a single start date for the treatment, so pre_post is just set to zero for all observations (in both treatment and control groups) that are dated before that date, and 1 for all observations that are after that date. In your case, you don't have a classic DID design and there are multiple treatment onset dates. Still for the treatment group it's easy: for each person, you set pre_post to zero for observations that precede that person's starting date and 1 for observations that follow it. The hard part is handling the control group. In #2 I suggested two approaches for setting the pre_post variable in the controls. Both of them entail identifying for each person in the control group a date at which they "would have started treatment" and then setting pre_post to 0 for observations before that date, and 1 for observations after that date.

Once you have those, the analysis proceeds by regressing the outcome on treatment group, pre_post, and their interaction. Just two dichotomous variables and their interaction. Now, sometimes you add covariates that are relevant, and there is always the issue of what type of regression, whether the outcome variable should be transformed in some way, etc. Those issues are always quite specific to the particular problem under study. But the core of DID is just those two variables and their interaction. In the end, the coefficient of the interaction is the DID estimator of the treatment effect.

[Sumgait Usagi]

Dear Clyde,

thank you for your comments. I truly appreciate for your time and suggestions. I will start to do that. However, couple of questions:

1) So you are suggesting is to create 32 pre_post variables and 32 interaction variables in total. So, Should I run all these treatment, pre-posts (32 var), and interaction (32) in the same regression?

2) if Q1 holds, then I may end-up in 32 different coefficients. I think some of them will be significant, and some not. Then how can i calculate the Aggregate ATE or whether overall program participation affected the outcome?

3) For comparison/control group, I would like to do it purely randomly. Then, Do you think it is better just to divide the dataset into several datasets? Then do it the calculations.

I welcome any opinion, recommendation to move further. I am stuck for now. Any STATA code that would ease my pressure would be also greatly helpful as well.
Thank you

[Clyde Schechter]

1) So you are suggesting is to create 32 pre_post variables and 32 interaction variables in total. So, Should I run all these treatment, pre-posts (32 var), and interaction (32) in the same regression?

No!!! What did I say that lead you to think that? I try to be clear in my writing here, and I don't understand how I failed so miserably. You need to create ONE pre-post variable and ONE treatment variable (and then one interaction term).

I am going to resist the temptation to show you sample code because it appears that we are presently far from having a common understanding of the problem.

Your question 2 is not relevant as it is premised on the incorrect assumption of 1), as already noted.

3) For comparison/control group, I would like to do it purely randomly. Then, Do you think it is better just to divide the dataset into several datasets? Then do it the calculations.

I do not understand what you are asking here. Divide the data set into several datasets based on what? What would be in each of the several data sets. What calculations would you do in each separate data set. Off hand, I can't think of any reason to divide up the data set for this kind of analysis, but as I clearly do not understand what your data are like and what you are doing, I can't definitely say you are wrong here.

[Sumgait Usagi]

Dear Clyde,

thank you for trying to explain. I appreciate it. Maybe I misunderstood you. I attached the part of the data to clear things up. some help is greatly appreciated.
There: "survey_date" - when people signed up to the survey; "survey" - whether they signed or not.
"date" overall consumption date, and "consumption" around people purchase for a given month.
"y=consumption"
"x= "treatment" "pre-post" "interaction" Individual characteristics (age, income, education, zipcode, ethnic group and etc)
(Overall, there are overall 43,000 obs (#3) for survey participants, about 200,000 non participants)

My treatment is "survey" - whether they signed up or not.
My treatment period - "pre_post" - "for each person, you set pre_post to zero for observations that precede that person's starting date and 1 for observations that follow it." (#4).
***My first confusion was how to create a single pre_post variable with 0 and 1, since I have multiple times of survey dates.
***How would that effect the control groups, and control group "pre_post" periods?
***Then issue to move forward to estimate the ATE?

Clearly there is a selection bias, so I want to employ following:
1) non-participants as comparison group (matching methods),
2) comparing early participants with late participants - to have better matched group.
Any suggestions for methods or coding? Am I clear a little bit? Again, I appreciate for your time and kind suggestions.
Regards.

[Clyde Schechter]

Please read the FAQ. Forum members are asked not to attach spreadsheets to provide example data. Some of the most frequent responders on this forum do not use Microsoft Office. Among those who do, some, including myself, will not download Office documents from a stranger because they can contain active malware. Everybody on this forum (presumably) uses Stata. The way to show examples of your data is to use the -dataex- command. Please read the entire FAQ, but you will find more about -dataex- in FAQ #12.

I'm going to assume your data contains the following variables: treatment (1 for treatment individuals, 0 for controls), start_date (the date at which people in the treatment group begin treatment), and current_date (the date of the observation.)

gen pre_post = (current_date > start_date) if !missing(current_date, start_date) if treatment == 1

will then set the pre_post variable in the treatment group. To set the values of the pre_post group in the controls you need to settle on a way to define a "date this person would have started treatment." As indicated in #2 this might be done by using some kind of predictive modeling, or it might be done by matching (on attributes or at random) each control to a treatment person and then using the treatment person's start date.

The coding of this kind of matching or predictive modeling is going to be more complicated than the code required to calculate variable pre_post itself or to run the regression once the data is set up. I sense that you are not yet prepared to undertake this. I urge you to step back from this project for a while and invest time learning the most basic Stata commands by reading the Getting Started [GS] and User's Guide[U] sections of the PDF manuals that came with your Stata installation. This is a fair amount of reading, and you won't retain everything. But it will acquaint you with the core commands that are used over and over in Stata data management and analysis. It will enable you, on seeing problems like this, to know at least in general which commands are likely to be useful in your particular situation and a rough idea of how to code them. Reference to the help files and corresponding sections in the user manuals will then guide you about the details. For example, familiarity with -gen- and -if- qualifiers from this reading would probably have enabled you to come up with the code shown above on your own. The time you spend doing this reading will be quickly repaid by decreased time puzzling over the basics, and less time posting questions here and waiting for somebody to answer them.

[Sumgait Usagi]

Dear Clyde,

thank you for your feedback. I already explained the data: There: "survey_date" - when people signed up to the survey; "survey" - whether they signed or not.
"date" - monthly consumption, and "consumption" amount people purchase for a given month.

Is this helpful for you or anyone wants to help? I have explained the problem in #1, and #7.
Thanks in advance.

* Example generated by -dataex-. To install: ssc install dataex
clear
input int ID long date float survey_date byte survey long consumption
 1 556   . 0  223
 1 557   . 0  864
 1 558   . 0  661
 1 559   . 0  794
 1 560   . 0  651
 1 561   . 0  242
 1 562   . 0  257
 1 563   . 0  353
 1 564   . 0  296
 1 565   . 0  424
 1 566   . 0  252
 1 567   . 0  270
 1 568   . 0  362
 1 569   . 0  444
 1 570   . 0  632
 1 571   . 0  531
 1 572   . 0  624
 1 573   . 0  447
 1 574   . 0  296
 1 575   . 0  267
 1 577   . 0  226
 1 578   . 0  160
 1 580   . 0  360
 1 581   . 0  810
 1 582   . 0  616
 1 583   . 0  585
 1 584   . 0  768
 1 585   . 0  287
 1 586   . 0  278
 1 587   . 0  305
 1 588   . 0  354
 1 589   . 0  306
 1 590   . 0  188
 1 590   . 0  115
 1 591   . 0  270
 1 592   . 0  426
 1 593   . 0  725
 1 594   . 0  663
 1 595   . 0  582
 1 596   . 0  623
 1 597   . 0  506
 1 598   . 0  333
 1 599   . 0  328
 1 600   . 0 1022
 1 601   . 0  598
 1 602   . 0  539
 1 603   . 0  277
 1 604   . 0  432
 1 605   . 0  645
 1 606   . 0  622
 1 607   . 0  714
 1 610   . 0  406
 1 611   . 0  543
 1 612   . 0  804
 1 613   . 0  659
 1 614   . 0  305
 1 615   . 0  283
 2 556 596 1  255
 2 557 596 1  476
 2 558 596 1  617
 2 559 596 1  649
 2 560 596 1  502
 2 561 596 1  289
 2 562 596 1  641
 2 563 596 1 1140
 2 564 596 1  869
 2 565 596 1 1512
 2 566 596 1  514
 2 568 596 1  225
 2 569 596 1  535
 2 570 596 1  547
 2 571 596 1  817
 2 573 596 1  412
 2 574 596 1  697
 2 575 596 1  586
 2 576 596 1 1144
 2 577 596 1 1197
 2 578 596 1  258
 2 579 596 1  201
 2 580 596 1  153
 2 581 596 1  699
 2 582 596 1  601
 2 583 596 1  607
 2 584 596 1  519
 2 585 596 1  228
 2 586 596 1  142
 2 586 596 1  546
 2 587 596 1 1262
 2 588 596 1  868
 2 589 596 1 1644
 2 590 596 1  774
 2 591 596 1  380
 2 592 596 1  398
 2 593 596 1  547
 2 594 596 1  593
 2 595 596 1  594
 2 596 596 1  775
 2 597 596 1  775
 2 598 596 1  441
 2 600 596 1 2255
 2 601 596 1 1549
 2 603 596 1  774
 2 604 596 1  478
 2 605 596 1  903
 2 606 596 1  761
 2 607 596 1  474
 2 608 596 1  401
 2 609 596 1  187
 2 610 596 1  409
 2 611 596 1  970
 2 612 596 1 1517
 2 613 596 1  816
 2 614 596 1  172
 2 615 596 1  171
 3 556   . 0   68
 3 557   . 0   80
 3 559   . 0 1217
 3 560   . 0 1273
 3 562   . 0  730
 3 563   . 0  622
 3 564   . 0  644
 3 565   . 0  647
 3 566   . 0  550
 3 567   . 0  726
 3 568   . 0  703
 3 569   . 0 1254
 3 570   . 0 1397
 3 571   . 0 1490
 3 572   . 0 1912
 3 573   . 0 1006
 3 573   . 0 1263
 3 575   . 0  724
 3 576   . 0  717
 3 577   . 0  618
 3 578   . 0  646
 3 579   . 0  546
 3 580   . 0  740
 3 581   . 0 1218
 3 582   . 0 1376
 3 583   . 0 1684
 3 584   . 0 1662
 3 585   . 0  745
 3 587   . 0  740
 3 587   . 0  651
 3 589   . 0  800
 3 590   . 0  653
 3 590   . 0  572
 3 591   . 0  652
 3 593   . 0 1226
 3 595   . 0 1526
 3 596   . 0 1079
 3 597   . 0 1157
 3 598   . 0  780
 3 599   . 0  457
 3 600   . 0  758
 3 601   . 0  750
 3 602   . 0  729
 3 603   . 0  650
 3 603   . 0  652
 3 605   . 0 1405
 3 605   . 0 1781
 3 607   . 0 1613
 3 607   . 0 2074
 3 609   . 0 1852
 3 610   . 0  753
 3 610   . 0  733
 3 611   . 0 1260
 3 613   . 0 1167
 3 614   . 0  722
 3 614   . 0  603
 3 616   . 0 1082
 4 556   . 0  384
 4 557   . 0  794
 4 558   . 0  773
 4 559   . 0  906
 4 560   . 0  665
 4 561   . 0  476
 4 562   . 0  411
 4 564   . 0  397
 4 565   . 0  397
 4 566   . 0  279
 4 567   . 0  290
 4 569   . 0  503
 4 570   . 0  653
 4 571   . 0  864
 4 572   . 0  622
 4 573   . 0  386
 4 574   . 0  273
 4 575   . 0  373
 4 576   . 0  505
 4 577   . 0  369
 4 578   . 0  364
 4 579   . 0  123
 4 579   . 0  161
 4 580   . 0  449
 4 581   . 0  633
 4 582   . 0  603
 4 583   . 0  571
 4 584   . 0  663
 4 585   . 0  299
 4 586   . 0  340
 4 587   . 0  403
 4 588   . 0  597
 4 589   . 0  412
 4 590   . 0  384
 4 591   . 0  396
 4 592   . 0  433
 4 593   . 0  704
 4 594   . 0  879
 4 595   . 0  623
 4 597   . 0  448
 4 599   . 0  497
 4 600   . 0  559
 4 601   . 0  502
 4 602   . 0  403
 4 603   . 0  365
 4 604   . 0  701
 4 605   . 0  636
 4 606   . 0  695
 4 607   . 0  763
 4 608   . 0  758
 4 609   . 0  458
 4 610   . 0  355
 4 611   . 0  481
 4 612   . 0  526
 4 613   . 0  473
 4 614   . 0  335
 4 615   . 0  460
 6 558   . 0 2208
 6 559   . 0 1965
 6 560   . 0 2148
 6 561   . 0 1327
 6 562   . 0  656
 6 563   . 0  766
 6 564   . 0  961
 6 565   . 0  766
 6 566   . 0  746
 6 567   . 0  733
 6 568   . 0  977
 6 569   . 0 1327
 6 570   . 0 2342
 6 571   . 0 2451
 6 572   . 0 2614
 6 573   . 0 1743
 6 574   . 0  734
 6 575   . 0  845
 6 576   . 0  872
 6 577   . 0  754
 6 578   . 0  750
 6 579   . 0  779
 6 580   . 0  799
 6 581   . 0 1661
 6 582   . 0 2096
 6 583   . 0 2201
 6 584   . 0 1616
 6 585   . 0 1083
 6 586   . 0  718
 6 588   . 0  433
 6 589   . 0  836
 6 590   . 0  706
 6 591   . 0  795
 6 592   . 0  793
 6 593   . 0 1213
 6 594   . 0 2239
 6 595   . 0 1715
 6 596   . 0 1711
 6 597   . 0 1332
 6 598   . 0  960
 6 599   . 0  778
 6 600   . 0  766
 6 601   . 0  690
 6 602   . 0  687
 6 603   . 0  731
 6 604   . 0  739
 6 605   . 0 1161
 6 606   . 0 1879
 6 607   . 0 2093
 6 608   . 0 1861
 6 609   . 0 1881
 6 610   . 0  783
 6 611   . 0  678
 6 612   . 0  866
 6 613   . 0  738
 6 614   . 0  703
 6 615   . 0  744
 6 616   . 0  890
 7 556 612 1  728
 7 557 612 1  963
 7 558 612 1 1616
 7 559 612 1 2096
 7 560 612 1 2038
 7 561 612 1  704
 7 563 612 1  499
 7 564 612 1  615
 7 565 612 1  568
 7 567 612 1  507
 7 568 612 1  748
 7 569 612 1 1188
 7 570 612 1 1794
 7 571 612 1 1725
 7 572 612 1 2025
 7 573 612 1 1145
 7 574 612 1  483
 7 575 612 1  500
 7 576 612 1  721
 7 577 612 1  543
 7 578 612 1  518
 7 579 612 1  499
 7 580 612 1  446
 7 581 612 1 1047
 7 582 612 1 1605
 7 583 612 1 1845
 7 584 612 1 1573
 7 585 612 1  618
 7 586 612 1  603
 7 587 612 1  607
 7 588 612 1  482
 7 588 612 1  693
 7 589 612 1  362
 7 590 612 1  553
 7 591 612 1  511
 7 592 612 1  561
 7 593 612 1 1027
 7 596 612 1 1172
 7 597 612 1  965
 7 598 612 1  700
 7 599 612 1 1031
 7 600 612 1 1189
 7 601 612 1  899
 7 602 612 1  805
 7 603 612 1  718
 7 604 612 1  565
 7 605 612 1 1144
 7 606 612 1 1668
 7 607 612 1 1915
 7 608 612 1 1586
 7 609 612 1 1271
 7 610 612 1  574
 7 611 612 1  523
 7 612 612 1  624
 7 613 612 1  504
 7 614 612 1  466
 7 615 612 1  406
 8 559   . 0  671
 8 560   . 0  830
 8 561   . 0  897
 8 562   . 0  340
 8 563   . 0  402
 8 564   . 0  447
 8 565   . 0  534
 8 566   . 0  419
 8 567   . 0  325
 8 568   . 0  437
 8 569   . 0  577
 8 570   . 0  864
 8 571   . 0  793
 8 572   . 0  898
 8 573   . 0  842
 8 574   . 0  391
 8 575   . 0  344
 8 576   . 0  841
 8 577   . 0  868
 8 578   . 0  633
 8 579   . 0  395
 8 580   . 0  420
 8 581   . 0  681
 8 582   . 0  778
 8 583   . 0  738
 8 584   . 0  781
 8 585   . 0  542
 8 586   . 0  331
 8 587   . 0  210
 8 588   . 0  386
 8 589   . 0  629
 8 591   . 0  349
 8 592   . 0  360
 8 593   . 0  549
 8 594   . 0 1093
 8 595   . 0  769
 8 596   . 0  826
 8 597   . 0  595
 8 598   . 0  411
 8 599   . 0  382
 8 600   . 0  535
 8 601   . 0  594
 8 602   . 0  543
 8 603   . 0  358
 8 604   . 0  369
 8 605   . 0  776
 8 606   . 0  940
 8 607   . 0 1146
 8 608   . 0  989
 8 609   . 0 1047
 8 610   . 0  546
 8 611   . 0  460
 8 612   . 0  809
 8 613   . 0  703
 8 614   . 0  440
 8 616   . 0  714
 9 556 587 1 1570
 9 557 587 1 2004
 9 558 587 1 2408
 9 559 587 1 2084
 9 560 587 1 2241
 9 561 587 1 1164
 9 562 587 1 1487
 9 563 587 1 2514
 9 564 587 1 1856
 9 565 587 1 2654
 9 566 587 1 1799
 9 567 587 1 1611
 9 568 587 1 4586
 9 569 587 1 3286
 9 570 587 1 2662
 9 571 587 1 2671
 9 572 587 1 2287
 9 573 587 1 1480
 9 574 587 1 1918
 9 575 587 1 4204
 9 576 587 1 4096
 9 578 587 1 1771
 9 580 587 1 1466
 9 581 587 1 1908
 9 582 587 1 1769
 9 583 587 1 2148
 9 584 587 1 2264
 9 585 587 1 1195
 9 586 587 1 1614
 9 587 587 1 3198
 9 588 587 1 1913
 9 588 587 1 1701
 9 589 587 1 2078
 9 590 587 1 2237
 9 591 587 1 1418
 9 592 587 1 1277
 9 593 587 1 1691
 9 594 587 1 2330
 9 595 587 1 1973
 9 596 587 1 1723
 9 597 587 1 1515
 9 598 587 1 1359
 9 599 587 1 2171
 9 600 587 1 5388
 9 601 587 1 3101
 9 602 587 1 4174
 9 603 587 1  699
 9 604 587 1  712
 9 606 587 1 1496
 9 607 587 1 1678
 9 608 587 1 1592
 9 609 587 1 1016
 9 610 587 1  750
 9 611 587 1  883
 9 613 587 1  758
 9 614 587 1  313
 9 615 587 1  325
10 560 610 1  594
10 561 610 1  848
10 562 610 1  739
10 563 610 1  939
10 564 610 1 1042
10 565 610 1 1066
10 566 610 1  743
10 567 610 1  750
10 568 610 1  818
10 569 610 1 1149
10 570 610 1 1030
10 572 610 1 1100
10 573 610 1  987
10 574 610 1  822
10 575 610 1  755
10 576 610 1  926
10 578 610 1  765
10 579 610 1  639
10 580 610 1  850
10 581 610 1  936
10 582 610 1  802
10 583 610 1 1164
10 584 610 1 1241
10 585 610 1 1035
10 586 610 1  872
10 587 610 1 1153
10 588 610 1  960
10 588 610 1  522
10 589 610 1  750
10 590 610 1  881
10 591 610 1  614
10 592 610 1  866
10 593 610 1  837
10 594 610 1 1170
10 595 610 1 1126
10 596 610 1 1087
10 597 610 1 1050
10 598 610 1  714
10 599 610 1  703
10 600 610 1 1362
10 601 610 1  898
10 602 610 1  918
10 604 610 1  858
10 605 610 1  962
end
format %tm date
format %tm survey_date
label values date date1
label var ID "group(phyaddr1)" 
label var survey "0 or 1"

[Clyde Schechter]

OK. So the first step now is to match controls with cases. As there is no variable in your data which could serve as a criterion for matching, we will do a random pairing.

* Example generated by -dataex-. To install: ssc install dataex
clear
input int ID long date float survey_date byte survey long consumption
 1 556   . 0  223
 1 557   . 0  864
 1 558   . 0  661
 1 559   . 0  794
 1 560   . 0  651
 1 561   . 0  242
 1 562   . 0  257
 1 563   . 0  353
 1 564   . 0  296
 1 565   . 0  424
 1 566   . 0  252
 1 567   . 0  270
 1 568   . 0  362
 1 569   . 0  444
 1 570   . 0  632
 1 571   . 0  531
 1 572   . 0  624
 1 573   . 0  447
 1 574   . 0  296
 1 575   . 0  267
 1 577   . 0  226
 1 578   . 0  160
 1 580   . 0  360
 1 581   . 0  810
 1 582   . 0  616
 1 583   . 0  585
 1 584   . 0  768
 1 585   . 0  287
 1 586   . 0  278
 1 587   . 0  305
 1 588   . 0  354
 1 589   . 0  306
 1 590   . 0  188
 1 590   . 0  115
 1 591   . 0  270
 1 592   . 0  426
 1 593   . 0  725
 1 594   . 0  663
 1 595   . 0  582
 1 596   . 0  623
 1 597   . 0  506
 1 598   . 0  333
 1 599   . 0  328
 1 600   . 0 1022
 1 601   . 0  598
 1 602   . 0  539
 1 603   . 0  277
 1 604   . 0  432
 1 605   . 0  645
 1 606   . 0  622
 1 607   . 0  714
 1 610   . 0  406
 1 611   . 0  543
 1 612   . 0  804
 1 613   . 0  659
 1 614   . 0  305
 1 615   . 0  283
 2 556 596 1  255
 2 557 596 1  476
 2 558 596 1  617
 2 559 596 1  649
 2 560 596 1  502
 2 561 596 1  289
 2 562 596 1  641
 2 563 596 1 1140
 2 564 596 1  869
 2 565 596 1 1512
 2 566 596 1  514
 2 568 596 1  225
 2 569 596 1  535
 2 570 596 1  547
 2 571 596 1  817
 2 573 596 1  412
 2 574 596 1  697
 2 575 596 1  586
 2 576 596 1 1144
 2 577 596 1 1197
 2 578 596 1  258
 2 579 596 1  201
 2 580 596 1  153
 2 581 596 1  699
 2 582 596 1  601
 2 583 596 1  607
 2 584 596 1  519
 2 585 596 1  228
 2 586 596 1  142
 2 586 596 1  546
 2 587 596 1 1262
 2 588 596 1  868
 2 589 596 1 1644
 2 590 596 1  774
 2 591 596 1  380
 2 592 596 1  398
 2 593 596 1  547
 2 594 596 1  593
 2 595 596 1  594
 2 596 596 1  775
 2 597 596 1  775
 2 598 596 1  441
 2 600 596 1 2255
 2 601 596 1 1549
 2 603 596 1  774
 2 604 596 1  478
 2 605 596 1  903
 2 606 596 1  761
 2 607 596 1  474
 2 608 596 1  401
 2 609 596 1  187
 2 610 596 1  409
 2 611 596 1  970
 2 612 596 1 1517
 2 613 596 1  816
 2 614 596 1  172
 2 615 596 1  171
 3 556   . 0   68
 3 557   . 0   80
 3 559   . 0 1217
 3 560   . 0 1273
 3 562   . 0  730
 3 563   . 0  622
 3 564   . 0  644
 3 565   . 0  647
 3 566   . 0  550
 3 567   . 0  726
 3 568   . 0  703
 3 569   . 0 1254
 3 570   . 0 1397
 3 571   . 0 1490
 3 572   . 0 1912
 3 573   . 0 1006
 3 573   . 0 1263
 3 575   . 0  724
 3 576   . 0  717
 3 577   . 0  618
 3 578   . 0  646
 3 579   . 0  546
 3 580   . 0  740
 3 581   . 0 1218
 3 582   . 0 1376
 3 583   . 0 1684
 3 584   . 0 1662
 3 585   . 0  745
 3 587   . 0  740
 3 587   . 0  651
 3 589   . 0  800
 3 590   . 0  653
 3 590   . 0  572
 3 591   . 0  652
 3 593   . 0 1226
 3 595   . 0 1526
 3 596   . 0 1079
 3 597   . 0 1157
 3 598   . 0  780
 3 599   . 0  457
 3 600   . 0  758
 3 601   . 0  750
 3 602   . 0  729
 3 603   . 0  650
 3 603   . 0  652
 3 605   . 0 1405
 3 605   . 0 1781
 3 607   . 0 1613
 3 607   . 0 2074
 3 609   . 0 1852
 3 610   . 0  753
 3 610   . 0  733
 3 611   . 0 1260
 3 613   . 0 1167
 3 614   . 0  722
 3 614   . 0  603
 3 616   . 0 1082
 4 556   . 0  384
 4 557   . 0  794
 4 558   . 0  773
 4 559   . 0  906
 4 560   . 0  665
 4 561   . 0  476
 4 562   . 0  411
 4 564   . 0  397
 4 565   . 0  397
 4 566   . 0  279
 4 567   . 0  290
 4 569   . 0  503
 4 570   . 0  653
 4 571   . 0  864
 4 572   . 0  622
 4 573   . 0  386
 4 574   . 0  273
 4 575   . 0  373
 4 576   . 0  505
 4 577   . 0  369
 4 578   . 0  364
 4 579   . 0  123
 4 579   . 0  161
 4 580   . 0  449
 4 581   . 0  633
 4 582   . 0  603
 4 583   . 0  571
 4 584   . 0  663
 4 585   . 0  299
 4 586   . 0  340
 4 587   . 0  403
 4 588   . 0  597
 4 589   . 0  412
 4 590   . 0  384
 4 591   . 0  396
 4 592   . 0  433
 4 593   . 0  704
 4 594   . 0  879
 4 595   . 0  623
 4 597   . 0  448
 4 599   . 0  497
 4 600   . 0  559
 4 601   . 0  502
 4 602   . 0  403
 4 603   . 0  365
 4 604   . 0  701
 4 605   . 0  636
 4 606   . 0  695
 4 607   . 0  763
 4 608   . 0  758
 4 609   . 0  458
 4 610   . 0  355
 4 611   . 0  481
 4 612   . 0  526
 4 613   . 0  473
 4 614   . 0  335
 4 615   . 0  460
 6 558   . 0 2208
 6 559   . 0 1965
 6 560   . 0 2148
 6 561   . 0 1327
 6 562   . 0  656
 6 563   . 0  766
 6 564   . 0  961
 6 565   . 0  766
 6 566   . 0  746
 6 567   . 0  733
 6 568   . 0  977
 6 569   . 0 1327
 6 570   . 0 2342
 6 571   . 0 2451
 6 572   . 0 2614
 6 573   . 0 1743
 6 574   . 0  734
 6 575   . 0  845
 6 576   . 0  872
 6 577   . 0  754
 6 578   . 0  750
 6 579   . 0  779
 6 580   . 0  799
 6 581   . 0 1661
 6 582   . 0 2096
 6 583   . 0 2201
 6 584   . 0 1616
 6 585   . 0 1083
 6 586   . 0  718
 6 588   . 0  433
 6 589   . 0  836
 6 590   . 0  706
 6 591   . 0  795
 6 592   . 0  793
 6 593   . 0 1213
 6 594   . 0 2239
 6 595   . 0 1715
 6 596   . 0 1711
 6 597   . 0 1332
 6 598   . 0  960
 6 599   . 0  778
 6 600   . 0  766
 6 601   . 0  690
 6 602   . 0  687
 6 603   . 0  731
 6 604   . 0  739
 6 605   . 0 1161
 6 606   . 0 1879
 6 607   . 0 2093
 6 608   . 0 1861
 6 609   . 0 1881
 6 610   . 0  783
 6 611   . 0  678
 6 612   . 0  866
 6 613   . 0  738
 6 614   . 0  703
 6 615   . 0  744
 6 616   . 0  890
 7 556 612 1  728
 7 557 612 1  963
 7 558 612 1 1616
 7 559 612 1 2096
 7 560 612 1 2038
 7 561 612 1  704
 7 563 612 1  499
 7 564 612 1  615
 7 565 612 1  568
 7 567 612 1  507
 7 568 612 1  748
 7 569 612 1 1188
 7 570 612 1 1794
 7 571 612 1 1725
 7 572 612 1 2025
 7 573 612 1 1145
 7 574 612 1  483
 7 575 612 1  500
 7 576 612 1  721
 7 577 612 1  543
 7 578 612 1  518
 7 579 612 1  499
 7 580 612 1  446
 7 581 612 1 1047
 7 582 612 1 1605
 7 583 612 1 1845
 7 584 612 1 1573
 7 585 612 1  618
 7 586 612 1  603
 7 587 612 1  607
 7 588 612 1  482
 7 588 612 1  693
 7 589 612 1  362
 7 590 612 1  553
 7 591 612 1  511
 7 592 612 1  561
 7 593 612 1 1027
 7 596 612 1 1172
 7 597 612 1  965
 7 598 612 1  700
 7 599 612 1 1031
 7 600 612 1 1189
 7 601 612 1  899
 7 602 612 1  805
 7 603 612 1  718
 7 604 612 1  565
 7 605 612 1 1144
 7 606 612 1 1668
 7 607 612 1 1915
 7 608 612 1 1586
 7 609 612 1 1271
 7 610 612 1  574
 7 611 612 1  523
 7 612 612 1  624
 7 613 612 1  504
 7 614 612 1  466
 7 615 612 1  406
 8 559   . 0  671
 8 560   . 0  830
 8 561   . 0  897
 8 562   . 0  340
 8 563   . 0  402
 8 564   . 0  447
 8 565   . 0  534
 8 566   . 0  419
 8 567   . 0  325
 8 568   . 0  437
 8 569   . 0  577
 8 570   . 0  864
 8 571   . 0  793
 8 572   . 0  898
 8 573   . 0  842
 8 574   . 0  391
 8 575   . 0  344
 8 576   . 0  841
 8 577   . 0  868
 8 578   . 0  633
 8 579   . 0  395
 8 580   . 0  420
 8 581   . 0  681
 8 582   . 0  778
 8 583   . 0  738
 8 584   . 0  781
 8 585   . 0  542
 8 586   . 0  331
 8 587   . 0  210
 8 588   . 0  386
 8 589   . 0  629
 8 591   . 0  349
 8 592   . 0  360
 8 593   . 0  549
 8 594   . 0 1093
 8 595   . 0  769
 8 596   . 0  826
 8 597   . 0  595
 8 598   . 0  411
 8 599   . 0  382
 8 600   . 0  535
 8 601   . 0  594
 8 602   . 0  543
 8 603   . 0  358
 8 604   . 0  369
 8 605   . 0  776
 8 606   . 0  940
 8 607   . 0 1146
 8 608   . 0  989
 8 609   . 0 1047
 8 610   . 0  546
 8 611   . 0  460
 8 612   . 0  809
 8 613   . 0  703
 8 614   . 0  440
 8 616   . 0  714
 9 556 587 1 1570
 9 557 587 1 2004
 9 558 587 1 2408
 9 559 587 1 2084
 9 560 587 1 2241
 9 561 587 1 1164
 9 562 587 1 1487
 9 563 587 1 2514
 9 564 587 1 1856
 9 565 587 1 2654
 9 566 587 1 1799
 9 567 587 1 1611
 9 568 587 1 4586
 9 569 587 1 3286
 9 570 587 1 2662
 9 571 587 1 2671
 9 572 587 1 2287
 9 573 587 1 1480
 9 574 587 1 1918
 9 575 587 1 4204
 9 576 587 1 4096
 9 578 587 1 1771
 9 580 587 1 1466
 9 581 587 1 1908
 9 582 587 1 1769
 9 583 587 1 2148
 9 584 587 1 2264
 9 585 587 1 1195
 9 586 587 1 1614
 9 587 587 1 3198
 9 588 587 1 1913
 9 588 587 1 1701
 9 589 587 1 2078
 9 590 587 1 2237
 9 591 587 1 1418
 9 592 587 1 1277
 9 593 587 1 1691
 9 594 587 1 2330
 9 595 587 1 1973
 9 596 587 1 1723
 9 597 587 1 1515
 9 598 587 1 1359
 9 599 587 1 2171
 9 600 587 1 5388
 9 601 587 1 3101
 9 602 587 1 4174
 9 603 587 1  699
 9 604 587 1  712
 9 606 587 1 1496
 9 607 587 1 1678
 9 608 587 1 1592
 9 609 587 1 1016
 9 610 587 1  750
 9 611 587 1  883
 9 613 587 1  758
 9 614 587 1  313
 9 615 587 1  325
10 560 610 1  594
10 561 610 1  848
10 562 610 1  739
10 563 610 1  939
10 564 610 1 1042
10 565 610 1 1066
10 566 610 1  743
10 567 610 1  750
10 568 610 1  818
10 569 610 1 1149
10 570 610 1 1030
10 572 610 1 1100
10 573 610 1  987
10 574 610 1  822
10 575 610 1  755
10 576 610 1  926
10 578 610 1  765
10 579 610 1  639
10 580 610 1  850
10 581 610 1  936
10 582 610 1  802
10 583 610 1 1164
10 584 610 1 1241
10 585 610 1 1035
10 586 610 1  872
10 587 610 1 1153
10 588 610 1  960
10 588 610 1  522
10 589 610 1  750
10 590 610 1  881
10 591 610 1  614
10 592 610 1  866
10 593 610 1  837
10 594 610 1 1170
10 595 610 1 1126
10 596 610 1 1087
10 597 610 1 1050
10 598 610 1  714
10 599 610 1  703
10 600 610 1 1362
10 601 610 1  898
10 602 610 1  918
10 604 610 1  858
10 605 610 1  962
end
format %tm date
format %tm survey_date
label values date date1
label var ID "group(phyaddr1)" 
label var survey "0 or 1"

//    SAVE THE DATA FOR LATER USE
tempfile holding
save `holding'

//    SPLIT OUT THE CONTROL GROUP IDs INTO A SEPARATE
//    FILE OF CONTROL IDS SORTED IN RANDOM ORDER
keep if survey == 0
keep ID
duplicates drop
set seed 1234
gen double shuffle = runiform()
sort shuffle
rename ID ID0
tempfile control_ids
save `control_ids'

//    NOW CREATE A FILE OF CASE IDs
use `holding', clear
keep if survey == 1
keep ID
rename ID ID1
duplicates drop

//    CREATE MATCHED PAIRS
merge 1:1 _n using `control_ids', keep(match) nogenerate
isid ID0
isid ID1, sort
gen long pair = _n

//    GO TO LONG LAYOUT AND BRING BACK THE OTHER VARIABLES
reshape long ID, i(pair) j(_j)
merge 1:m ID using `holding', keep(match) nogenerate
assert survey == _j
drop _j

//    PREPARE PRE-POST VARIABLE
//    BEGIN BY APPLYING CASE SURVEY DATE TO MATCHED CONTROL
by pair (survey_date), sort: assert survey_date == survey_date[1] if !missing(survey_date)
by pair (survey_date): replace survey_date = survey_date[1]
gen byte pre_post = (date >= survey_date)

//    DID ANALYSIS
xtset ID
xtreg consumption i.survey##i.pre_post, fe vce(cluster pair)

Notes

1. Your example data is not panel data: there are multiple observations of the same ID and date with different values of consumption. Do check your actual data for this. It isn't necessarily a problem, but it's unusual to see this and it suggests that your data set may contain errors.

2. Because you have more controls than cases, you lose a few of the controls in the matching.

[Sumgait Usagi]

Dear Clyde,

thank you very much for your suggestions and notes. I will work on it. Keep you posted. I am sure this would be useful for other statalist members and community.

Kind regards.

[Sumgait Usagi]

Dear Clyde,

after the reshape, I keep getting the following with the full sample. I tried to drop any duplicates.
// GO TO LONG LAYOUT AND BRING BACK THE OTHER VARIABLES reshape long ID, i(pair) j(_j) no xij variables found You typed something like reshape wide a b, i(i) j(j). reshape looked for existing variables named a# and b# but could not find any. Remember this picture: long wide +---------------+ +------------------+ | i j a b | | i a1 a2 b1 b2 | |---------------| <--- reshape ---> |------------------| | 1 1 1 2 | | 1 1 3 2 4 | | 1 2 3 4 | | 2 5 7 6 8 | | 2 1 5 6 | +------------------+ | 2 2 7 8 | +---------------+ long to wide: reshape wide a b, i(i) j(j) (j existing variable) wide to long: reshape long a b, i(i) j(j) (j new variable) Any reason, why would this be the case. I use the sample data I share with you, I don't get the above. Thank you very much.

[Sumgait Usagi]

Sorry for the above:
after the reshape, I keep getting the following with the full sample:
reshape long ID, i(pair) j(_j)

Any reason, why would this be the case. I use the sample data I share with you, I don't get the above. Thank you very much.

[Clyde Schechter]

The code worked perfectly with the example data you gave previously. If you changed the name of the ID variable or if you made changes to the code that used a different variable name, that could account for this problem. In any case, in order to troubleshoot this, I would need you to post a new example of data that produces this problem when run with that code.

[Sumgait Usagi]

Dear Clyde,

Problem I raised above happens when I add more observations or when I add new variable to the data (even with small data, adding new variable creates the problem).
(recap: Estimate ATE of survey participation, given that different households signs up surveys at different dates. Data also include non-survey participants, as well, pre and post periods of the consumption
"y=consumption"
"x= "treatment/survey" Individual characteristics (age, income, education, zipcode, ethnic group and etc))

* Example generated by -dataex-. To install: ssc install dataex
clear
input int ID long date float survey_date byte survey long(consumption con_purchase)
 1 564   . 0  296   6
 1 565   . 0  428   7
 1 566   . 0  252   4
 1 567   . 0  270   3
 1 568   . 0  362   3
 1 569   . 0  333   6
 1 570   . 0  332   9
 1 571   . 0  531  11
 1 572   . 0  624  11
 1 573   . 0  447   9
 1 574   . 0  296   9
 1 575   . 0  267   4
 1 577   . 0  226   6
 1 578   . 0  160   7
 1 580   . 0  360   9
 1 581   . 0  610   9
 1 582   . 0  610  16
 1 583   . 0  580  14
 1 584   . 0  268   8
 1 585   . 0  287   0
 1 586   . 0  278   0
 1 587   . 0  305   7
 1 588   . 0  354   7
 1 589   . 0  306   8
 1 590   . 0  303   8
 1 591   . 0  270   8
 1 592   . 0  426  11
 1 593   . 0  725   9
 1 594   . 0  263   9
 1 595   . 0  582  11
 1 596   . 0  423  11
 1 597   . 0  306  11
 1 598   . 0  222   8
 1 599   . 0  228   8
 1 600   . 0  622   5
 1 601   . 0  598   8
 1 602   . 0  539   9
 1 603   . 0  317   2
 1 604   . 0  432   2
 1 605   . 0  245  22
 1 606   . 0  422   6
 1 607   . 0  217   6
 1 610   . 0  406   6
 1 611   . 0  543   6
 1 612   . 0  804   7
 1 613   . 0  659   7
 1 614   . 0  305   7
 1 615   . 0  283   7
 2 556 596 1  255   6
 2 557 596 1  476   7
 2 558 596 1  617  16
 2 559 596 1  649   8
 2 560 596 1  502   9
 2 561 596 1  289   8
 2 562 596 1  641  11
 2 563 596 1  840  10
 2 564 596 1  869   7
 2 565 596 1  780   8
 2 566 596 1  514   4
 2 568 596 1  225   8
 2 569 596 1  535   7
 2 570 596 1  547   7
 2 571 596 1  817   7
 2 573 596 1  412   0
 2 574 596 1  697   7
 2 575 596 1  586   6
 2 576 596 1  644   0
 2 577 596 1  697   7
 2 578 596 1  258   2
 2 579 596 1  201  11
 2 580 596 1  153   8
 2 581 596 1  466  19
 2 582 596 1  301  16
 2 583 596 1  407   8
 2 584 596 1  519   7
 2 585 596 1  228  11
 2 586 596 1  688   6
 2 587 596 1  862   3
 2 588 596 1  868   6
 2 589 596 1  644   5
 2 590 596 1  774   2
 2 591 596 1  380   5
 2 592 596 1  398   8
 2 593 596 1  547   5
 2 594 596 1  593   6
 2 595 596 1  594   7
 2 596 596 1  775   5
 2 597 596 1  775   6
 2 598 596 1  441   6
 2 600 596 1 1255   4
 2 601 596 1 1349   0
 2 603 596 1  774   0
 2 604 596 1  478   3
 2 605 596 1  703   4
 2 606 596 1  761   2
 2 607 596 1  474   2
 2 608 596 1  401   2
 2 609 596 1  187   0
 2 610 596 1  409   4
 2 611 596 1  970   2
 2 612 596 1 1157   3
 2 613 596 1  816   1
 2 614 596 1  172   7
 2 615 596 1  171   0
 3 556   . 0   68   0
 3 557   . 0   80  15
 3 559   . 0 1027  11
 3 560   . 0 1273  11
 3 562   . 0  730  18
 3 563   . 0  622  15
 3 564   . 0  644  15
 3 565   . 0  647  19
 3 566   . 0  550  13
 3 567   . 0  726  17
 3 568   . 0  703   0
 3 569   . 0 1254  17
 3 570   . 0 1397  17
 3 571   . 0 1490  19
 3 572   . 0 1712  19
 3 573   . 0 1163   0
 3 574   . 0 1006   0
 3 575   . 0  724   9
 3 576   . 0  717   9
 3 577   . 0  618  14
 3 578   . 0  646  14
 3 579   . 0  546  18
 3 580   . 0  740  19
 3 581   . 0 1218  18
 3 582   . 0 1376  18
 3 583   . 0 1684  18
 3 584   . 0 1662  11
 3 585   . 0  745   8
 3 586   . 0  651   0
 3 587   . 0  740   0
 3 588   . 0  800   8
 3 589   . 0  653  19
 3 590   . 0  572   0
 3 591   . 0  652  11
 3 593   . 0 1226  11
 3 595   . 0 1156  21
 3 596   . 0 1079  25
 3 597   . 0  807  27
 3 598   . 0  780  11
 3 599   . 0  457  11
 3 600   . 0  758  17
 3 601   . 0  450  17
 3 602   . 0  429  19
 3 603   . 0  450   0
 3 604   . 0  652   0
 3 605   . 0 1571   0
 3 606   . 0 1405  11
 3 607   . 0 1163   0
 3 608   . 0 1074  11
 3 609   . 0 1852   0
 3 610   . 0  753  19
 3 611   . 0  733   0
 3 612   . 0 1160   0
 3 613   . 0  867   0
 3 614   . 0  722  11
 3 615   . 0  603   0
 3 616   . 0  882 101
 4 556   . 0  284  24
 4 557   . 0  694  75
 4 558   . 0  773  79
 4 559   . 0  906  15
 4 560   . 0  665  15
 4 561   . 0  476  12
 4 562   . 0  411  22
 4 564   . 0  397  14
 4 565   . 0  397  15
 4 566   . 0  279  17
 4 567   . 0  290  17
 4 569   . 0  503  18
 4 570   . 0  653  13
 4 571   . 0  864  14
 4 572   . 0  622   9
 4 573   . 0  386  12
 4 574   . 0  273  15
 4 575   . 0  373  14
 4 576   . 0  505   0
 4 577   . 0  369   0
 4 578   . 0  364   7
 4 579   . 0  284   3
 4 580   . 0  449   3
 4 581   . 0  633  13
 4 582   . 0  603  11
 4 583   . 0  571  11
 4 584   . 0  663   0
 4 585   . 0  299   0
 4 586   . 0  340  18
 4 587   . 0  403  17
 4 588   . 0  597  23
 4 589   . 0  412  14
 4 590   . 0  384  35
 4 591   . 0  396  15
 4 592   . 0  433  15
 4 593   . 0  704   0
 4 594   . 0  879  11
 4 595   . 0  623   7
 4 597   . 0  448  18
 4 599   . 0  497  16
 4 600   . 0  559  12
 4 601   . 0  502  11
 4 602   . 0  403  12
 4 603   . 0  365   0
 4 604   . 0  701  31
 4 605   . 0  636  18
 4 606   . 0  695   0
 4 607   . 0  763   9
 4 608   . 0  758  17
 4 609   . 0  458  17
 4 610   . 0  355   0
 4 611   . 0  481   8
 4 612   . 0  526   9
 4 613   . 0  473  13
 4 614   . 0  335  18
 4 615   . 0  460  19
 6 558   . 0 2208  15
 6 559   . 0 1965  16
 6 560   . 0 2148  19
 6 561   . 0 1327   7
 6 562   . 0  656  17
 6 563   . 0  766   6
 6 564   . 0  961  23
 6 565   . 0  766  14
 6 566   . 0  746  12
 6 567   . 0  733   0
 6 568   . 0  977   0
 6 569   . 0 1327   0
 6 570   . 0 2342  16
 6 571   . 0 2451   7
 6 572   . 0 2614   3
 6 573   . 0 1743   7
 6 574   . 0  734  25
 6 575   . 0  845  18
 6 576   . 0  872  16
 6 577   . 0  754  11
 6 578   . 0  750  10
 6 579   . 0  779  10
 6 580   . 0  799  12
 6 581   . 0 1661  29
 6 582   . 0 2096  13
 6 583   . 0 2201  18
 6 584   . 0 1616   9
 6 585   . 0 1083  17
 6 586   . 0  718  10
 6 588   . 0  433   0
 6 589   . 0  836   0
 6 590   . 0  706   6
 6 591   . 0  795  15
 6 592   . 0  793   0
 6 593   . 0 1213   9
 6 594   . 0 2239  10
 6 595   . 0 1715   0
 6 596   . 0 1711 126
 6 597   . 0 1332   0
 6 598   . 0  960   0
 6 599   . 0  778   0
 6 600   . 0  766  24
 6 601   . 0  690  39
 6 602   . 0  687   8
 6 603   . 0  731  12
 6 604   . 0  739  11
 6 605   . 0 1161   0
 6 606   . 0 1879  15
 6 607   . 0 2093  11
 6 608   . 0 1861   9
 6 609   . 0 1881   5
 6 610   . 0  783  14
 6 611   . 0  678   8
 6 612   . 0  866  19
 6 613   . 0  738   0
 6 614   . 0  703  14
 6 615   . 0  744   0
 6 616   . 0  890  15
 7 556 612 1  728   1
 7 557 612 1  963   4
 7 558 612 1 1616   8
 7 559 612 1 2096   8
 7 560 612 1 2038   6
 7 561 612 1  704   5
 7 563 612 1  499   4
 7 564 612 1  615   2
 7 565 612 1  568   3
 7 567 612 1  507   8
 7 568 612 1  748   2
 7 569 612 1 1188   7
 7 570 612 1 1794  13
 7 571 612 1 1725   8
 7 572 612 1 2025   7
 7 573 612 1 1145  33
 7 574 612 1  483  29
 7 575 612 1  500   0
 7 576 612 1  721   0
 7 577 612 1  543  13
 7 578 612 1  518  19
 7 579 612 1  499  31
 7 580 612 1  446  22
 7 581 612 1 1047  15
 7 582 612 1 1605  21
 7 583 612 1 1845   7
 7 584 612 1 1573   1
 7 585 612 1  618  10
 7 586 612 1  603   8
 7 587 612 1  607   5
 7 588 612 1  693   9
 7 589 612 1  362   0
 7 590 612 1  553   5
 7 591 612 1  511   7
 7 592 612 1  561   0
 7 593 612 1 1027  10
 7 596 612 1 1172   4
 7 597 612 1  965   0
 7 598 612 1  700   8
 7 599 612 1 1031  11
 7 600 612 1 1189   5
 7 601 612 1  899   5
 7 602 612 1  805  24
 7 603 612 1  718  33
 7 604 612 1  565  11
 7 605 612 1 1144   0
 7 606 612 1 1668  17
 7 607 612 1 1915  14
 7 608 612 1 1586  21
 7 609 612 1 1271   0
 7 610 612 1  574   7
 7 611 612 1  523   0
 7 612 612 1  624  18
 7 613 612 1  504   0
 7 614 612 1  466  17
 7 615 612 1  406   0
 8 559   . 0  671   9
 8 560   . 0  830  19
 8 561   . 0  897  37
 8 562   . 0  340  19
 8 563   . 0  402  22
 8 564   . 0  447  23
 8 565   . 0  534  20
 8 566   . 0  419  20
 8 567   . 0  325  22
 8 568   . 0  437  22
 8 569   . 0  577  23
 8 570   . 0  864  25
 8 571   . 0  793  21
 8 572   . 0  898  24
 8 573   . 0  842  23
 8 574   . 0  391  22
 8 575   . 0  344  19
 8 576   . 0  841  22
 8 577   . 0  868   0
 8 578   . 0  633  33
 8 579   . 0  395  18
 8 580   . 0  420   0
 8 581   . 0  681  25
 8 582   . 0  778  38
 8 583   . 0  738  28
 8 584   . 0  781  40
 8 585   . 0  542  41
 8 586   . 0  331  32
 8 587   . 0  210  33
 8 588   . 0  386  43
 8 589   . 0  629   0
 8 591   . 0  349  31
 8 592   . 0  360  43
 8 593   . 0  549  39
 8 594   . 0 1093   0
 8 595   . 0  769  39
 8 596   . 0  826   0
 8 597   . 0  595  41
 8 598   . 0  411   0
 8 599   . 0  382  71
 8 600   . 0  535  56
 8 601   . 0  594  45
 8 602   . 0  543  39
 8 603   . 0  358  43
 8 604   . 0  369   0
 8 605   . 0  776   0
 8 606   . 0  940  33
 8 607   . 0 1146   0
 8 608   . 0  989  72
 8 609   . 0 1047  47
 8 610   . 0  546  73
 8 611   . 0  460  71
 8 612   . 0  809  74
 8 613   . 0  703   0
 8 614   . 0  440  65
 8 616   . 0  714  44
 9 556 587 1 1570   0
 9 557 587 1 2004   0
 9 558 587 1 2408  93
 9 559 587 1 2084  39
 9 560 587 1 2241  63
 9 561 587 1 1164  74
 9 562 587 1 1487  62
 9 563 587 1 2514 100
 9 564 587 1 1856  51
 9 565 587 1 2654  49
 9 566 587 1 1799  86
 9 567 587 1 1611  67
 9 568 587 1 4586   0
 9 569 587 1 3286 110
 9 570 587 1 2662  92
 9 571 587 1 2671  85
 9 572 587 1 2287  81
 9 573 587 1 1480  65
 9 574 587 1 1918  77
 9 575 587 1 4204 105
 9 576 587 1 4096 262
 9 578 587 1 1771  44
 9 580 587 1 1466  78
 9 581 587 1 1908  73
 9 582 587 1 1769  46
 9 583 587 1 2148  68
 9 584 587 1 2264  78
 9 585 587 1 1195  55
 9 586 587 1 1614  69
 9 587 587 1 3198   0
 9 588 587 1 3614  51
 9 589 587 1 2078  72
 9 590 587 1 2237  48
 9 591 587 1 1418  61
 9 592 587 1 1277   5
 9 593 587 1 1691   0
 9 594 587 1 2330   0
 9 595 587 1 1973  65
 9 596 587 1 1723  55
 9 597 587 1 1515   0
 9 598 587 1 1359  54
 9 599 587 1 2171  56
 9 600 587 1 5388 407
 9 601 587 1 3101  79
 9 602 587 1 4174  89
 9 603 587 1  699  84
 9 604 587 1  712  88
 9 606 587 1 1496  75
 9 607 587 1 1678  81
 9 608 587 1 1592  85
 9 609 587 1 1016   0
 9 610 587 1  750  68
 9 611 587 1  883  68
 9 613 587 1  758  66
 9 614 587 1  313  39
 9 615 587 1  325  34
10 560 610 1  594  17
10 561 610 1  848  49
10 562 610 1  739   0
10 563 610 1  939  46
10 564 610 1 1042  49
10 565 610 1 1066  36
10 566 610 1  743  39
10 567 610 1  750  53
10 568 610 1  818  38
10 569 610 1 1149  52
10 570 610 1 1030  44
10 572 610 1 1100  42
10 573 610 1  987  40
10 574 610 1  822  51
10 575 610 1  755  45
10 576 610 1  926  45
10 578 610 1  765  44
10 579 610 1  639  39
10 580 610 1  850  51
10 581 610 1  936  43
10 582 610 1  802  27
10 583 610 1 1164  54
10 584 610 1 1241  56
10 585 610 1 1035  61
10 586 610 1  872  61
10 587 610 1 1153  69
10 588 610 1 1488  66
10 589 610 1  750  78
10 590 610 1  881  20
10 591 610 1  614  47
10 592 610 1  866  60
10 593 610 1  837   0
10 594 610 1 1170  49
10 595 610 1 1126   0
10 596 610 1 1087  55
10 597 610 1 1050  53
10 598 610 1  714  48
10 599 610 1  703  43
10 600 610 1 1362  63
10 601 610 1  898  40
10 602 610 1  918  49
10 604 610 1  858  53
10 605 610 1  962  45
10 606 610 1 1047  52
10 607 610 1 1221  52
10 608 610 1 1235   0
10 610 610 1  961  65
10 611 610 1  812  53
10 612 610 1 1294  70
10 613 610 1 1223  59
10 614 610 1  834  58
10 615 610 1  639  48
11 556   . 0  776  55
11 557   . 0 1287  73
11 558   . 0 1205   0
11 559   . 0 1215   0
11 560   . 0 1112  50
11 561   . 0  722   0
11 562   . 0  837  49
11 563   . 0  964  56
11 564   . 0  924  47
11 565   . 0 1154  47
11 566   . 0  593  75
11 567   . 0  584  95
11 568   . 0  952   0
11 569   . 0 1242  86
11 570   . 0 1147  70
11 571   . 0 1416   0
11 572   . 0 1022  54
11 573   . 0  786   0
11 574   . 0  817  70
11 575   . 0  728  50
11 576   . 0 1127   0
11 577   . 0  916  48
11 578   . 0  752  67
11 579   . 0  637  58
11 581   . 0 1263  49
11 582   . 0 1142  48
11 583   . 0  939  34
11 584   . 0  960  41
11 585   . 0  560   0
11 586   . 0  584  38
11 588   . 0  824  44
11 589   . 0 1008  49
11 590   . 0  529  47
11 591   . 0  426   0
11 592   . 0  535  63
11 593   . 0 1004  75
11 594   . 0 1130  54
11 595   . 0  644  58
11 596   . 0 1431  46
11 597   . 0  844  44
11 598   . 0  530  44
11 599   . 0  872  44
11 600   . 0 1694  94
11 601   . 0  885  33
11 602   . 0  789  40
11 603   . 0  578  54
11 604   . 0  826  58
11 605   . 0 1362  60
11 606   . 0 1246  34
11 607   . 0 1430  37
11 608   . 0 1348  45
11 609   . 0  697  34
11 610   . 0  549  37
11 611   . 0 1055  43
11 612   . 0 1349  35
11 614   . 0  413  31
11 615   . 0  614  28
12 556 600 1  640  28
12 557 600 1 1286   0
12 558 600 1 1451  29
12 559 600 1 1716  39
12 560 600 1 1368  62
12 561 600 1  654  69
12 562 600 1  297  18
12 563 600 1  430  31
12 564 600 1  337  23
12 565 600 1  370  24
12 566 600 1  356  24
12 567 600 1  346  23
12 568 600 1  744  21
12 569 600 1 1422  17
12 570 600 1 1641   0
12 571 600 1 1201  13
12 572 600 1 1048  21
12 573 600 1  677  17
12 574 600 1  434  31
12 575 600 1  416  27
12 577 600 1  186  26
12 578 600 1  226  17
12 579 600 1  282  22
12 580 600 1  780  37
12 582 600 1 1487   0
12 583 600 1 1478  31
12 584 600 1 1544  47
12 585 600 1  454  37
12 586 600 1  235  35
12 587 600 1  281  47
12 588 600 1  278  65
12 589 600 1  261   0
12 590 600 1  270  43
12 591 600 1  256  17
12 592 600 1  639  43
12 593 600 1 1583  58
12 594 600 1 1758  46
12 595 600 1 1893  47
12 596 600 1  959  35
12 597 600 1  722  33
12 598 600 1  277  36
12 599 600 1  251  41
12 600 600 1  308  72
12 601 600 1  251  71
12 602 600 1  237  93
12 603 600 1  241  85
12 604 600 1  498  32
12 605 600 1 1532  43
12 606 600 1 1319  36
12 607 600 1 1639  42
12 609 600 1  467  38
12 611 600 1  518  43
12 612 600 1  413  52
12 613 600 1  241  33
12 614 600 1  230  38
12 615 600 1  301  44
13 556 593 1 2010 102
13 557 593 1 3356 106
13 558 593 1 2982  94
13 559 593 1 3138 107
13 560 593 1 2793  97
13 561 593 1 1692   0
13 562 593 1 1382   0
13 563 593 1 1562  83
13 564 593 1 1604  93
13 565 593 1 1940 144
13 566 593 1 1644  87
13 567 593 1 1795  87
13 568 593 1 2467 162
13 569 593 1 3120 152
13 570 593 1 2963  95
13 571 593 1 3332  99
13 572 593 1 3003  93
13 573 593 1 2329 148
13 574 593 1 1737 186
13 575 593 1 1630  76
13 576 593 1 1746  81
13 577 593 1 1475  68
13 578 593 1 1541  84
13 579 593 1 1380  82
13 580 593 1 2215  92
13 581 593 1 2896  94
13 582 593 1 2633  87
13 583 593 1 2393  72
13 584 593 1 2747  83
13 586 593 1 1379  77
13 588 593 1 1921  93
13 589 593 1 1696  85
13 590 593 1 1602  80
13 591 593 1 1477  42
13 592 593 1 1830   0
13 593 593 1 2929   0
13 594 593 1 2710 104
13 595 593 1 2441  75
13 596 593 1 2169  78
13 597 593 1 1551  65
13 598 593 1 1113  68
13 599 593 1 1046   0
13 600 593 1 1437  88
13 601 593 1 1151  62
13 602 593 1 1087  62
13 603 593 1 1007  66
13 604 593 1 1338  79
13 605 593 1 2595  84
13 606 593 1 2230  66
13 607 593 1 2773  89
13 608 593 1 2612  97
13 609 593 1 1711  95
13 610 593 1 1251  82
13 611 593 1 1241   0
13 612 593 1 1559  83
13 613 593 1 1449   0
13 614 593 1  947   0
15 556   . 0  858  40
15 557   . 0 1303  50
15 558   . 0 1179  50
15 559   . 0 1463  42
15 560   . 0 1106  35
15 561   . 0  857  35
15 562   . 0  907   0
15 563   . 0  809   0
15 564   . 0  989  62
15 565   . 0  914  33
15 566   . 0  749  71
15 567   . 0  732  53
15 568   . 0  828 163
15 569   . 0 1262  36
15 570   . 0 1228  29
15 571   . 0 1196  31
15 572   . 0  893  28
15 573   . 0  838  29
15 574   . 0  939  31
15 575   . 0  583  27
15 576   . 0  631  47
15 577   . 0  719  48
15 578   . 0  597  44
15 579   . 0  552  35
15 582   . 0  986  49
15 583   . 0 1189  45
15 584   . 0  979  40
15 585   . 0  767  43
15 586   . 0  668  37
15 587   . 0  665  39
15 588   . 0  735  44
15 589   . 0  413  37
15 590   . 0  508  39
15 591   . 0  514  11
15 592   . 0  684   0
15 593   . 0 1235   0
15 594   . 0  984 112
15 595   . 0 1116   0
15 596   . 0 1023  42
15 597   . 0  724  42
15 598   . 0  659  41
15 599   . 0  718  45
15 600   . 0 1322  43
15 601   . 0  585  37
15 602   . 0  550  55
15 603   . 0  530  61
15 604   . 0  798  72
15 605   . 0 1137  75
15 606   . 0 1083 105
15 607   . 0 1163 139
15 609   . 0  601  55
15 610   . 0  531  27
15 611   . 0  536  32
15 612   . 0  581   0
15 613   . 0  405  16
15 615   . 0  423  27
16 556   . 0  662  47
16 557   . 0  989  62
16 558   . 0 1197  69
16 559   . 0 1095  53
16 560   . 0  943   0
16 561   . 0  782  63
16 562   . 0  530   0
16 563   . 0  697  55
16 564   . 0  802  43
16 565   . 0  626  43
16 566   . 0  602  47
16 567   . 0  633 159
16 568   . 0  764   0
16 569   . 0  919  55
16 570   . 0 1170  63
16 571   . 0 1144  47
16 572   . 0  930  38
16 573   . 0  880  39
16 574   . 0  706  46
16 575   . 0  674  46
16 577   . 0  663  38
16 578   . 0  638  37
16 579   . 0  639  40
16 580   . 0  798  45
16 582   . 0 1159  47
16 583   . 0 1131  52
16 584   . 0  965  32
16 585   . 0  817  44
16 586   . 0  665  27
16 587   . 0  689  32
16 588   . 0  803   0
16 589   . 0  778  46
16 590   . 0  641  41
16 591   . 0  707  40
16 592   . 0  775  35
16 593   . 0  960   0
16 594   . 0 1296  47
16 595   . 0 1042  47
16 596   . 0  929  49
16 597   . 0  792  26
16 598   . 0  572  28
16 599   . 0  633  32
16 600   . 0  780  36
16 601   . 0  599  32
16 602   . 0  628  57
16 603   . 0  785   1
16 604   . 0  810   0
16 605   . 0  969 151
16 606   . 0  938  39
16 607   . 0 1192  29
16 609   . 0  755  40
16 610   . 0  744  39
16 611   . 0  629  43
16 613   . 0  626  27
16 614   . 0  575  38
16 615   . 0  686   0
17 556   . 0 1427  16
17 557   . 0 2442  32
17 558   . 0 2605  50
17 559   . 0 1783  59
17 560   . 0 3394  47
17 561   . 0 1551  45
17 562   . 0 1288  42
17 563   . 0 1815  54
17 564   . 0 1553  48
17 565   . 0 1976  35
17 566   . 0 1286  37
17 567   . 0 1381  45
17 568   . 0 1687  37
17 569   . 0 2091  37
17 570   . 0 2343  25
17 571   . 0 2592  45
17 572   . 0 2278  35
17 573   . 0 1383   0
17 574   . 0 1364  27
17 575   . 0 1557  26
17 576   . 0 2046  38
17 577   . 0 1910   0
17 579   . 0 1387  46
17 580   . 0 1544  31
17 581   . 0 2139  23
17 582   . 0 2053  22
17 584   . 0 2125  23
17 585   . 0 1056  19
17 586   . 0 1260  22
17 587   . 0 1822  26
17 588   . 0 2167  29
17 589   . 0 1314  29
17 590   . 0 1396  19
17 591   . 0 1006  13
17 592   . 0 1436  22
17 593   . 0 2243  37
17 594   . 0 2084  26
17 595   . 0 2235  46
17 596   . 0 2291  78
17 597   . 0 1983  88
17 598   . 0 1471  62
17 599   . 0 2037   0
17 600   . 0 3323  79
17 601   . 0 2390  36
17 602   . 0 2030 110
17 603   . 0 1878 162
17 604   . 0 2243  98
17 605   . 0 3355   0
17 606   . 0 2671   0
17 607   . 0 2713  55
17 608   . 0 1859  24
17 609   . 0 1141  29
17 610   . 0 1216  25
17 613   . 0 2628   0
17 614   . 0 1530   0
17 615   . 0 1453   0
21 556   . 0  790 195
21 557   . 0 1169 229
21 559   . 0 1721 265
21 561   . 0 1171 210
21 562   . 0 1225 221
21 563   . 0 1508   0
21 564   . 0 1282 244
21 565   . 0 1638 154
21 566   . 0  862   0
21 567   . 0  829 115
21 568   . 0 1056 272
21 569   . 0 1407 169
21 570   . 0 1528 135
21 571   . 0 1696   0
21 572   . 0 1623 152
21 573   . 0 1554 119
21 574   . 0 1924 128
21 575   . 0 1867 132
21 576   . 0 2060 154
21 577   . 0 1616 121
21 578   . 0 1168 119
21 579   . 0 1148   0
21 580   . 0 1212 183
21 581   . 0 1517 276
21 582   . 0 1516 265
21 583   . 0 1309 122
21 584   . 0 1523 192
21 585   . 0  831  65
21 586   . 0 1067  75
21 587   . 0 1615  52
21 588   . 0  318   0
21 589   . 0 1371  23
21 590   . 0 1282 362
21 592   . 0  890 124
21 593   . 0 1421 244
21 594   . 0 1498   0
21 595   . 0 1050 219
21 596   . 0 1207  31
21 597   . 0 1254   0
21 598   . 0 1009 101
21 599   . 0 1334   0
21 600   . 0 2523 315
21 601   . 0 1711 124
21 602   . 0 1437  86
21 603   . 0 1087 185
21 604   . 0 1018 161
21 605   . 0 1446  85
21 606   . 0 1417 122
21 607   . 0 1684 172
21 608   . 0 1636 138
21 609   . 0 1005 513
21 610   . 0 1153  56
21 611   . 0 2069   0
21 612   . 0 2731  83
21 613   . 0 2071   0
21 614   . 0  999   0
21 615   . 0  986   0
22 556   . 0 1232  60
22 557   . 0 1833  74
22 558   . 0 1705  57
22 559   . 0 1901  66
22 560   . 0 1698  67
22 562   . 0 1288  67
22 563   . 0 1634  72
22 564   . 0 1433  55
22 565   . 0 2124  65
22 567   . 0 1161   0
22 568   . 0 1477  63
22 569   . 0 1737  63
22 570   . 0 1533  47
22 571   . 0 1670  46
22 572   . 0 1459   0
22 573   . 0 1035  40
22 574   . 0 1399  52
22 575   . 0 1405  51
22 576   . 0 1689  52
22 577   . 0 1592  49
22 578   . 0 1295  61
22 579   . 0 1110  57
22 580   . 0 1033  45
22 581   . 0 1410  48
22 582   . 0 1352  35
22 583   . 0 1211  40
22 584   . 0 1470  53
22 585   . 0  884  48
22 586   . 0 1217  49
22 587   . 0 1909  49
22 588   . 0 2545  63
22 589   . 0 1435  47
22 590   . 0 1578  53
22 591   . 0  973  25
22 593   . 0 1424  44
22 594   . 0 1342  42
22 595   . 0 1320  49
22 596   . 0 1206  46
22 598   . 0 1034  44
22 599   . 0 1725   0
22 600   . 0 3072  57
22 601   . 0 2198  40
22 602   . 0 1643   0
22 603   . 0  910  45
22 604   . 0 1017  38
22 605   . 0 1492  57
22 606   . 0 1319  37
22 607   . 0 1789  56
22 608   . 0 1671  56
22 609   . 0 1022  55
22 610   . 0 1226  50
22 611   . 0 2183  57
22 612   . 0 2857  62
22 613   . 0 2185  50
22 614   . 0 1068  48
22 615   . 0 1153  58
23 556   . 0    0   0
23 557   . 0    1   0
23 558   . 0    0   0
23 559   . 0    0   0
23 561   . 0    0   0
23 562   . 0    0   0
23 563   . 0    0   0
23 564   . 0    0   0
23 565   . 0    0   0
23 566   . 0    0   0
23 567   . 0    1   0
23 568   . 0    0   0
23 569   . 0    0   0
23 570   . 0   14   0
23 571   . 0   33  10
23 572   . 0   34   1
23 573   . 0   38   0
23 574   . 0   56   0
23 575   . 0   59   0
23 576   . 0   63   0
23 577   . 0   70   0
23 578   . 0   59   0
23 579   . 0   54   0
23 580   . 0   54   0
23 581   . 0   40   0
23 582   . 0   43   0
23 583   . 0   42   0
23 584   . 0   40   0
23 585   . 0   44   0
23 586   . 0   81   0
23 587   . 0   91   0
23 588   . 0   28   0
23 589   . 0   71   0
23 591   . 0   51   0
23 592   . 0   44   0
23 593   . 0   41   0
23 594   . 0   36   0
23 595   . 0   37   0
23 596   . 0   37   0
23 597   . 0   43   0
23 598   . 0   50   0
23 599   . 0   60   0
23 600   . 0   77   4
23 601   . 0   61   2
23 602   . 0   60   3
23 603   . 0   53   4
23 604   . 0   46   2
23 605   . 0   37   2
23 606   . 0   20   1
23 607   . 0    0   1
23 608   . 0    0   0
23 609   . 0    0   1
23 610   . 0    0   0
23 611   . 0    0   1
end
format %tm date
format %tm survey_date
label values date date1