Dear Stata users,
I want to study the effects of protests on populism. I have a balanced panel and the following election-years: 1995, 2002, 2007, 2012, 2017, 2022. All treated municipalities were treated at the same time in 2022, 704 treated units, 33497 not-treated units. I want to estimate the effect in 2022.
What I did:
Results. Coefficient: 1.583 s.e. 0.151
I obtain the following figure:
Looking to the figure in the right-side, it seems that I have parallel trends. However, if I do:
it does not seem that I have parallel trends at all. I understand that xtdidregress and this last regression are doing different things, but which results are more relevant for testing parallel trends?
I want to study the effects of protests on populism. I have a balanced panel and the following election-years: 1995, 2002, 2007, 2012, 2017, 2022. All treated municipalities were treated at the same time in 2022, 704 treated units, 33497 not-treated units. I want to estimate the effect in 2022.
What I did:
Code:
xtdidregress (populism)(protest), group(municipality) time(year) estat ptrends estat trendplots
I obtain the following figure:
Looking to the figure in the right-side, it seems that I have parallel trends. However, if I do:
Code:
reg populism treated##ibn.year if year<2022, vce(cluster municipality) hascons
Linear regression Number of obs = 171,005
F(10, 34200) = 36914.58
Prob > F = 0.0000
R-squared = 0.2854
Root MSE = 4.5094
(Std. err. adjusted for 34,201 clusters in municipality)
------------------------------------------------------------------------------
| Robust
populism | Coefficient std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
1.treated | .5693092 .1393846 4.08 0.000 .2961107 .8425078
|
year |
1995 | 10.79962 .0278197 388.20 0.000 10.74509 10.85415
2002 | 11.68494 .0233048 501.40 0.000 11.63926 11.73062
2007 | 6.820931 .0169692 401.96 0.000 6.787671 6.854191
2012 | 10.72906 .0236022 454.58 0.000 10.6828 10.77533
2017 | 15.7514 .0301276 522.82 0.000 15.69235 15.81045
|
treated#year |
1 1995 | -.2740706 .1340293 -2.04 0.041 -.5367725 -.0113687
1 2002 | -.9777467 .120084 -8.14 0.000 -1.213115 -.742378
1 2007 | -.5563054 .1234835 -4.51 0.000 -.7983371 -.3142737
1 2012 | -.8389415 .0809102 -10.37 0.000 -.9975281 -.6803549
1 2017 | 0 (omitted)
------------------------------------------------------------------------------
