General Cluster Question

[Lindsey Buck]

Hi all,

I am new to STATA so this is a very basic question. I want to do a probit regression with controls for gender and state and year fixed effects. I need to cluster at the state*year of census level.
Would this look like:
"probit y variableofinterest gender i.statecode i.yearcode, cluster(state*year)"
Can you assist with this?
This isn't working but I am not sure how to run this code exactly.
Thanks!

[FernandoRios]

Hi Lindsay
The problem is that Stata does not understand what do you mean with Stata*Year, because it expects a single variable to be used there.
What you can do is the following:

egen state_year=group(stata year)
* or
gen long state_year=year*1000+state
* The second one assumes State is coded between 0 to 999.

After that you can use the new variable as a cluster
HTH