# Re: st: Estimating total factor productivity

 From "Scott Merryman" To statalist@hsphsun2.harvard.edu Subject Re: st: Estimating total factor productivity Date Tue, 8 Jul 2008 12:47:46 -0500

```On Mon, Jul 7, 2008 at 9:39 AM, dadhinp@yahoo.com <dadhinp@yahoo.com> wrote:
> Dear STATA users
> I want to estimate TFP. Since I have very small size of time series data set (20 obervation) I am using CD production function with CRS

TFP = Total Factor Productivity?
CD = Cobb-Douglas?
CRS= Constant Returns to Scale?

> and perfectly competetive market assumption. But I dont know how to write stata command. Following is the data set.
>
> Year     Ln(K/L)      Ln(Y/L)
>
>  1981     2.1715      8.4583
> 1982     2.1887      8.5757
> 1983     2.2061      8.5374
> 1984     2.2214      8.6244
> 1985     2.2346      8.7500
> 1986     2.2453      8.7886
> 1987     2.2558      8.8013
> 1988     2.2652      8.8624
> 1989     2.2736      8.9083
> 1990     2.2805      8.9494
> 1991     2.2875      9.0049
> 1992     2.2904      9.0085
> 1993     2.2934      8.9990
> 1994     2.2963      9.0333
> 1995     2.2989      9.0199
> 1996     2.3031      9.0336
> 1997     2.3070      9.0385
> 1998     2.3110      9.0307
> 1999     2.3138      9.0327
> 2000     2.3164      9.0534
>
> K=capital L=labor Ln=log
> I found one paper using the same data set estimated Ln(Y/L)=4.43 +0.46Log(K/L)

I doubt it.  Log(Y/L) is about 4 times greater than Log(K/L).  The
above regression would generate predicted values of around 5.

> But I get very different result. So could you please to suggest me STATA code to estimate CD function and to find TFP and Change in >TFP.

Perhaps this may be of use:

http://www.stata.com/statalist/archive/2003-12/msg00581.html

Scott
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```