hi,
I am estimating dynamic panel model(n=50, t=15).
My question is about the range of the estimate of the first differenced independent variable of dependent variable.
The rule of thumb by Bond(2001) said that the pooled ols estimate should be considered an upper bound estimate,
while the correspond fixed effect should a lower bound estimate.
But my estimate(0.92) is above the pooled estimate(0.88). It is estimated by system gmm(orthogonal option).
Is it reasonable? if not reasonable, please let me know any solution?
Thanks in advance.