A question about confidence intervals in stata

[John Williamss]

Are Stata ols and panel regression confidence intervals symmetric around the estimated value with standard options for any data? Specifically for these commands:
reg y x i.dummy , robust cluster(grp)

[Leonardo Guizzetti]

Yes, they are, which you could also easily verify. Please also note the use of old syntax for -regress-. This is the modern way of requesting cluster robust variance estimates.

regress y ...., vce(cluster clusterid)

But I have to ask, why does it matter if the CI is symmetric or not?

[Joro Kolev]

Yes, they are symmetric. What Stata reports are Wald confidence intervals, and these are symmetric by construction.