Dear all,
I'm estimating TFP at firm-year level. I have data for each firm at each year. I'm using Compustat data. I use perpetual method for capital and EMP for labor. I assume time-invariant labor and capital elasticity and loop through each sector. Thus I got time-invariant sector specific labor and capital elasticity. And then I use firm-year specific data to calculate TFP so I got firm-year specific TFP which changes across time and sector. However, some of my labor/capital elasticity are negative (the mean makes sense, 2/3 for labor, 1/3 for capital, but the min is negative), thus also some of my TFP estimate, does this make sense?
Thanks a lot!
I'm estimating TFP at firm-year level. I have data for each firm at each year. I'm using Compustat data. I use perpetual method for capital and EMP for labor. I assume time-invariant labor and capital elasticity and loop through each sector. Thus I got time-invariant sector specific labor and capital elasticity. And then I use firm-year specific data to calculate TFP so I got firm-year specific TFP which changes across time and sector. However, some of my labor/capital elasticity are negative (the mean makes sense, 2/3 for labor, 1/3 for capital, but the min is negative), thus also some of my TFP estimate, does this make sense?
Thanks a lot!
