Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

# Re: st: Stochastic frontier model_ constraining coefficients

 From Scott Merryman To statalist@hsphsun2.harvard.edu Subject Re: st: Stochastic frontier model_ constraining coefficients Date Fri, 17 Sep 2010 04:43:01 -0500

```In a translog production function, returns to scale involves more then
just the coefficients on labor and capital.

If the production function is written out as:

Y = b0 + b1Labor + b2Capital + b3Time + .5[b4Capital^2 + b5Labor^2 + b6Time^2] +
b7Capital*Labor + b8Capital*Time + b9Labor*Time + u_it - v_it

where all output capital and labor are in logs.

The elasticity of output with respect to capital and labor are:

e_k = b2 + b4Capital + b7Labor + b8Time

e_l = b1 + b5Labor + b7Capital + b9Time

e_k + e_l = Returns to Scale.  The translog production function may
exhibit increasing returns for some input combinations and decreasing
returns for others.

Scott

On Thu, Sep 16, 2010 at 3:03 PM,  <Declan.Curran@dcu.ie> wrote:
> Hello, I hope you can help me:
>
> I am running a time varying decay (panel) stochastic frontier model for a
> translog production function. The model converges but my capital and labour
> coefficients are -0.52 and 1.52, respectively (see output below). So, they
> add to 1 (constant returns) but they are not between 0-1.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```