Clustering standard error when the number of cluster are 11

[Tariku Getaneh]

Dear All,
I am conducting a diff-in-diff regression where households in 5 states were affected by a policy after 2005 and households in 6 states were not. I defined treatment at State level.
My model:
xtreg depVar treat*PostPol i.time, fe cluster(state) Is this sufficient? if not what can I do better?

Best regards,
TG

[Carlo Lazzaro]

Tariku:
11 clusters are too few.
I'd consider -bootstrap- standard errors.

[Tariku Getaneh]

Thank you Carlo. How do I practically conduct bootstrap? My knowledge is truly limited in this area. I really appreciate your insight/suggestion.
When you say, "11 clusters are too few" does this mean the SE's are unaccountable?

Best regards