Hello. I am using LP to estimate the TFP for firm level data as follows:
levpet ln_output, free( ln_lbr ) proxy( ln_materials) capital( ln_k) reps(2)
where output is the deflated sales, lbr is labor, materials and k are deflated materials value and capital value.
After this, I run:
predict tfp , omega
My question is whether i need to log the tfp using:
gen ln_tfp = ln(tfp)
when I sum the tfp variable, i get large values so i am assuming the tfp is not in logs currently using the technique so it has to be logged manually? Thank you!
levpet ln_output, free( ln_lbr ) proxy( ln_materials) capital( ln_k) reps(2)
where output is the deflated sales, lbr is labor, materials and k are deflated materials value and capital value.
After this, I run:
predict tfp , omega
My question is whether i need to log the tfp using:
gen ln_tfp = ln(tfp)
when I sum the tfp variable, i get large values so i am assuming the tfp is not in logs currently using the technique so it has to be logged manually? Thank you!
