Dear Stata users,
I am writing a paper in which I estimated a general (pooled) Tobit model. The dependent variable is left-censored at zero. In order to obtain parameter estimates and cluster-robust standard errors I ran the following line:
I want to replicate the cluster-robust standard errors manually, i.e., without the vce(cluster id) option.
For example, the estimated variance-covariance matrix from the standard Tobit, i.e.,
is easily replicated by the inverse of negative estimated Hessian from the (Tobit) log-likelihood function.
Any help is highly appreciated.
I am writing a paper in which I estimated a general (pooled) Tobit model. The dependent variable is left-censored at zero. In order to obtain parameter estimates and cluster-robust standard errors I ran the following line:
Code:
tobit pmedalshare lngdp lnpop polity host $yeardummies, ll(0) vce(cluster id)
For example, the estimated variance-covariance matrix from the standard Tobit, i.e.,
Code:
tobit pmedalshare lngdp lnpop polity host $yeardummies, ll(0)
Any help is highly appreciated.
Comment