Hello,

I need your help. I hope you are staying safe. I have an unbalanced panel dataset. It includes n=84 and t=9.
1) When we interpret the coefficient of time dummies, is there any difference for the dependent variable if it is "the rate of something" or "the level of something"? for example: growth rate or income per capita. Can we interpret both of them?
2) If time dummies are dropped in system GMM. How one can interpret?
For the first and the second question for an example i shared my results (see below)
3) Is there any "rules" in GMM to interpret results of time dummies?
4) Is there any "rule of thumb" to decide about "how may lag of dependent variable or iv, we should add to model."

Thank you.

The code which i executed : xtabond2 FKO l.FKO ARG l.ARG l2.ARG KO NSB ADH yr*, gmmstyle(l.FKO l.ARG l2.ARG, collapse) ivstyle(KO ADH NSB yr*, equation(level)) nodiffsargan twostep robust orthogonal small

Array