How to cluster nothing in a code?

[Phuc Nguyen]

I have a code that is

did2s depvar [if] [in] [weight], first_stage(varlist) second_stage(varlist) treatment(varname) cluster(varname)

But I do not want to cluster, so what I should do to run the code above without cluster?

When I ignore the cluster, it says

option cluster() required

[Rich Goldstein]

as the help file for that command tells you, "(use unit id if you don't want to cluster)"; see

help did2s

[Joro Kolev]

And what the help file says is generally true everywhere.

Clustered standard errors/variances with clustering at the unit level are equivalent to robust standard errors/variances.

[FernandoRios]

Also, based on the authors paper (Gardner 2021), and the recent work by Borusyak et al(2021), you have to cluster at least at the panel FE level.
If you look into the recent literature on DID models, you will find the relevant references

[Phuc Nguyen]

Also, based on the authors paper (Gardner 2021), and the recent work by Borusyak et al(2021), you have to cluster at least at the panel FE level.
If you look into the recent literature on DID models, you will find the relevant references

"you have to cluster at least at the panel FE level" do you mean the unit level then?

[FernandoRios]

You may want to read Borusyak et al(2021) paper. They explain that you need to cluster standard errors at least at the level of the unit FE you are absorbing. I provided the link to some of the information on your other post