Running regression on panel data using a single year of data

[Sonia Nassar]

Hi everyone,

The data that I'm using is from a study by Nicholas Bloom, Mirko Draca and John Van Reenen--the dataset includes annual information on patent applications for 115 German industries between 1996 and 2002. What I'm supposed to do is run a regression of log patent applications on Chinese import penetration for 2002 but I'm not quite sure how to approach this or how to take the log of the patent variable after reshaping.

This is the code that I've used so far:

use "$direc/German patent applications 1996.DTA", clear
append using "$direc/German patent applications 1997.DTA"
append using "$direc/German patent applications 1998.DTA"
append using "$direc/German patent applications 1999.DTA"
append using "$direc/German patent applications 2000.DTA"
append using "$direc/German patent applications 2001.DTA"
append using "$direc/German patent applications 2002.DTA"
xtset industry year
reshape wide patent china_imp, i(industry) j(year)

And this is what my data looks like:

Thanks for your help!