Imposing monotonicity on translog production function

[Pedro Soares]

Hi everyone,

I am running a stochastic frontier model but I need to impose monotonicity conditions on a translog prodution function. Does anyone know how I can do this in Stata?

Thank you.

[Abdul Adam]

Pedro. You would increase your chance of getting help if you show the forum an example of what you have done so far with the help of dataex (SSC before 15.1).
Having said that, I suggest you try:
help nlcom.

[Pedro Soares]

Abdul,

Thank you for the suggestions. Unfortunately I can't show an example because I'm using confidential microdata. But let me try to be more clear: I'm running a stochastic frontier model, using Kumbhakar's command sfmodel, but when I check monotonicity conditions of each input some observations violate this condition. The problem is, as showed by Sauer (Journal of Applied Economics 9: 139-165, 2006), when I have violations of monotonicity my estimates of technical efficiency are not consistents.
One way to solve this problem can be found in Henningsen and Henning (J Prod Anal 32: 217-229, 2009) that proposed a three-step procedure to impose monotonicity. But these authors used R commands to do this.
As showed by Henningsen and Henning (2009) what I need is use quadratic programming to calculate
\beta^0=argmin(\beta^0-\beta)*\Sigma_{\beta}^-1*(\beta^0-\beta)
s.t. f_{i}(x, \beta^0)>=0
where \beta are the parameters of the production frontier, \Sigma is the covariance matrix, \beta^0 the restricted parameters and f_{i} the partial derivative with respect to input i.
From what I saw, the only way to do this in Stata is using Mata, but this language is a problem for me. So if someone knows an alternative way to do this without Mata, I'd really appreciate!