Hello, would someone have any help on how to run a two limit tobit model with robust and clustered standard errors. I want to cluster on region (I also include region dummies i.region as fixed effects). My data N=~360,000 individuals nested within 310 regions with two years of data, 2000 and 2010 (not the same individuals across years). I have a count dependent variable with a range of values 0 to 30 (the number of days a person has experienced poor health over the past month). Having a lower limit of 0 (comprising about 64% of the observations) and an upper limit of 30 (about 5-6% of obs.). Thank you very much for any help.