[Mateus Maciel]

I could solve the issue by running the following:

update all, force

A special thanks to Pei-Chun from the Stata support and Fernando.

[Kagiso Matswalela]

Dear FernandoRios

After running the jwdid command, variables such as _etr_ and _tr_are created. How do we interpret these?

[FernandoRios]

those are the "internally" created variables. They are used to create the event-like study estimates. and to identify the effectively treated observations. You do not need to interpret them.

[Kagiso Matswalela]

Thank you

[Mustafa Caliskan]

Hi FernandoRios. Is there any option like CSDID long or long2 to avoid pretrend problems in JWDID?

[FernandoRios]

by default, if you use "never" as control group, results will be similar to long2
And, if you have pre-trend problems, there is no way to avoid them. they simply exist

[Ian Wong]

I also met the varlist required error after running estat event. I replaced the .ado files of the jwdid package with the two .ado files (jwdid.ado and jwdid_estat.ado) downloaded from FernandoRios 's website. https://friosavila.github.io/playing...ain_jwdid.html
The issue was solved. I installed my stata packages in D:\stata\ado\plus. By default, the packages can be in C:\....stata\ado\plus.

[Ian Wong]

oh, use the ado files here. I found that I should also replace jwdid_plot.ado file, or it couldn't work well either.
https://github.com/friosavila/stpack...ree/main/jwdid

[FernandoRios]

Thanks Ian for mentioning that.
I have yet to send the update to SSC
F
Note that is a bug only affecting Stata 17 users

[Ridwan Sheikh]

Dear FernandoRios

I am implementing JWDID in gravity setting. I have one minor clarification about the treatment assignment.

In a regional trade agreements database, If a country-pair "AB" first became members of RTA in 2005, and therefore may belong to g=2005 cohort. However, in the dataset of trade flows, when we merge it with RTA dataset, we observe that for given pair AB, the outcome (trade-flows) is availabe only for pre-treatment years, i.,e for years < 2005, we do not have data on outcome for this pair in post-treatment periods, i.e., for years > 2005.
I want to ask, should we put this pair in 2005 cohort or keep it in untreated group (control).
In other words, for pairs with the data on outcome available only in some years in the pre-treatment periods, and missing in post-periods when the pair is treated, should such country-pairs be included in the control group or in the treatmet-cohort when the pairs are treated.

Thanks,
(Ridwan)