Dear Stata-users
I have a panel in which unobserved time-invariant individual specific effects are drawn from a non-normal distribution, say from an exponential distribution with rate 0.5.
These effects are uncorrelated with both the idiosyncratic error terms and the covariates, so the key RE assumption of no correlation between the unobserved heterogeneity and the covariates is satisfied. The response variable is generated from a Tobit model.
My questions are as follows:
Does pooled tobit (using tobit) yield consistent estimates corresponding to the time-varying covariates?
Does the xttobit command provide consistent estimates corresponding to the time-varying covariates?
I have a panel in which unobserved time-invariant individual specific effects are drawn from a non-normal distribution, say from an exponential distribution with rate 0.5.
These effects are uncorrelated with both the idiosyncratic error terms and the covariates, so the key RE assumption of no correlation between the unobserved heterogeneity and the covariates is satisfied. The response variable is generated from a Tobit model.
My questions are as follows:
Does pooled tobit (using tobit) yield consistent estimates corresponding to the time-varying covariates?
Does the xttobit command provide consistent estimates corresponding to the time-varying covariates?
