d(log) log interpretation

[Daveoves]

Hi all,

I am having trouble interpreting my panel estimation

I have the following output dlog(y)= 13.905.log(x1) - 2.097.log(x2) + e

How do I interpret the coefficients ?

is it: a 1% increase in x1 will yield 13.905% increase in y ?

I am not sure if this is correct since its is the difference of log(y)... do you have to multiply or divide it by a 100 to get the same level of % change?

I am desperate

please help me

Thank

David

[dakila]

you have to take cumulative sum to get level of % change.