"prodest" stata command to estimate TFP

[Alberto Alvarez]

Dear Stata Users,

I want to use "prodest" command to estimate TFP. Can you, please adress the following concerns:

1. I want to follow methodology of Wooldridge (2009). Do I need to specify "gmm" below:

prodest logVA, free(logL) proxy(logInput) state(logK) method(wrdg) gmm i(firm) t(year) ?


2. The command provides residuals (TFP) in levels. Is it possible to get the residuals in real value?

[Sven-Kristjan Bormann]

To answer your questions:
1. You can specify the "gmm" option, but it is not necessary to run the Wooldridge methodology. As far as I understand it, the only thing that changes, is that instead of using -ivregress-, now -gmm- is used to estimate the equations.
2. I don't understand what you mean with TFP in real value. The TFP should be usually predicted as log-TFP. You can get the "normal" TFP via using -predict tfp, exponential-.