If we have the following regression based on the Solow model:
log(yi) = β0 + β1 * log(si) + β2 * log(ni + gi + δi ) + ei
And we know that based on:
y* = 〖(s/(n + δ + g))〗^(α/(1-α))
log(y*) = α/(1-α) * log(s) – α/(1-α) * log(n + δ + g) )
So we can assume that beta 1 and beta 2 are equal to: α/(1-α)
Of course beta 1 should be positive and beta 2 should be negative, but what if they have different values, for example beta 1 equals 1.5 and beta 2 equals -2.5?
What does this mean economically? I am a little bit confused about how beta 1 and beta 2 can be different.
Thank you very much for every answer!