4 components stochastic frontier analysis with heteroscedasticity

[M. Yun]

I am trying to estimate a stochastic frontier production function using panel data in the context of 4 components model with heteroscedasticity as in Lai and Kumbhakar 2018. They estimate the standard 4 componets model y_it = Bx_it + tau_i - eta_i + v_it - u_it, based on MSL approach.
However, I am not sure how to implement this procedure using STATA. Can anyone tell me what STATA command should be used to implement this estimation?

[Andrew Musau]

You can estimate the four- error components model (Generalized True Random Effects [GTRE]) in multiple steps, as we have done here. As far as I know, there are no implementations of the single-step maximum likelihood or simulated maximum likelihood estimators in Stata to date. If you send me an email, I can forward you the do-files from our paper in case you are interested in the multi-step estimation procedure.

[M. Yun]

Thank you so much for the info. I am interested in the multi-step estimation procedure. I sent you an email, and would appreciate it if you forwarded the ado file. Thanks again!

[Andrew Musau]

I didn't receive your email. Please note that I have changed institutions since that paper was published. Kindly use my current work email. If you have trouble finding it, feel free to send me a private message here.