Hi everyone, I have the following individual level county specific data where the treatment timing (impact_var) is different for each county. I looked into all other staggered diff in diff works and see that the treatment-control groups are mostly the ones that are treated vs ones never treated. However, my treatment control group is person with a car vs without a car. Since, this is not a balanced panel data, how should i go ahead and work on this? I tried to use bacon decompostion for dynamic diff-in-diff since TWFE would be biased.
Code:
* Example generated by -dataex-. For more info, type help dataex clear input byte p_id str1 county byte(month impact_month) str4 treatment_group double(outcome_var independent_var) 1 "A" 1 3 "yes" .2818808067933959 .5450594150323939 2 "A" 1 3 "yes " .9858339825468566 .4364299307352397 3 "A" 1 3 "no" .36669659003416266 .38070007081074986 4 "A" 2 3 "yes" .9599754006051351 .23211071912605896 5 "A" 2 3 "no" .7269755119894432 .19319421084816424 6 "A" 2 3 "yes" .07200209144556657 .1220095844198803 7 "A" 3 3 "yes " .5222588470002206 .04927778451415121 8 "A" 3 3 "no" .938889658355819 .756989449156155 9 "A" 3 3 "yes" .42793845007400344 .19642315958868828 10 "A" 4 3 "no" .19945284541894082 .7335298074692761 11 "A" 4 3 "yes" .22535759725413773 .26281918916732794 12 "A" 4 3 "yes " .5587509514416281 .0266742110158662 13 "A" 4 3 "no" .3658689030561856 .006432138747697436 14 "A" 5 3 "yes" .17700557681618667 .5603036207919811 15 "A" 5 3 "no" .10175744793298713 .18897232268140918 16 "B" 1 4 "yes" .6827773649754435 .953212011491134 17 "B" 2 4 "yes " .5197893392463795 .6664544382994466 18 "B" 2 4 "no" .5665108450502471 .7879986394463855 19 "B" 2 4 "yes" .5961963389772857 .7756572374966352 20 "B" 3 4 "no" .6016621668323542 .28828214954403764 21 "B" 3 4 "yes" .9974930757877027 .021347942986916446 22 "B" 3 4 "yes " .770227190975668 .1251575702794766 23 "B" 3 4 "no" .8908138051083169 .9102775602146674 24 "B" 4 4 "yes" .383870890932669 .0035538227297660097 25 "B" 4 4 "no" .9302803614124064 .9931037783949731 26 "B" 5 4 "yes" .6576135117831666 .2897146688657347 27 "B" 5 4 "yes " .6256040852611076 .9142197891452521 28 "B" 5 4 "no" .565560541049187 .21480041898494262 29 "C" 1 2 "yes" .5567823401919891 .6970189692640844 30 "C" 1 2 "no" .36619044171717163 .5423322552225732 31 "C" 1 2 "yes" .7604414752187605 .49058875645844036 32 "C" 2 2 "yes " .7615769687477215 .2287253051380156 33 "C" 2 2 "no" .39419862186235843 .3002463105850113 34 "C" 2 2 "yes" .03766309563771841 .6900612586102638 35 "C" 3 2 "no" .7568439386219061 .5244691113833923 36 "C" 3 2 "yes" .25950729400661476 .3301759292372449 37 "C" 3 2 "yes " .919434161989217 .6193096109447961 38 "C" 3 2 "no" .0011197131419192763 .5076571923195543 39 "C" 4 2 "yes" .5118936905829388 .612023659479903 40 "D" 1 3 "no" .07318737222495586 .6710930223872101 41 "D" 1 3 "yes" .2760449116068029 .03876267026640723 42 "D" 2 3 "yes " .33175932522160057 .4551023584127033 43 "D" 2 3 "no" .6937891021905837 .04280899860249976 44 "D" 3 3 "yes" .5057506478105401 .6048876852207012 45 "D" 3 3 "no" .43011710709497797 .95899535062922 46 "D" 3 3 "yes" .5864184475254749 .3887734936559192 47 "D" 3 3 "yes " .7972303146475893 .25916563551309313 48 "D" 4 3 "no" .3466640603313965 .42782844617665405 49 "D" 5 3 "yes" .4617208076611288 .47396866184440867 50 "D" 6 3 "no" .8865065672187957 .7471843894867372 end
