difference in difference with cem matching and clustering

[Veeresh Thummadi]

Hi Statalist users

I have a panel data with project panel observations. However, each project has owners and the owners are owners can have multiple projects. I wanted to cluster on the owner. In the normal regression I can cluster on owner. But if I try to do a cem matching with clustering, however, I am not able to use this option. Here is the command I ran. Can someone suggest how to overcome this challenge?

 regress performance i.treat##i.post## covariates [iweight=cem_weights], cluster(owner)
iweights not allowed with robust
r(101);

Best
Veeresh

[Gupta Manish]

Try: [aweight=cem_weights]

[Veeresh Thummadi]

Thanks Gupta Manish it worked