I have some questions concerning the function pgmm from the plm package. I want to control for reversed causality and thus, have to use a lagged model with an interaction effect. For that, I can use the pgmm function, but I am not sure how exactly my formula has to look like. So I want to regress y on x and on x*y(t-1), which is the effect of y on x . Is the following formula correct for this:
form = y~x+x:lag(y,1) | lag(x, 2:9)
And also, my data is not complete. That means, not all individuals are present in all time periods. How can I handle such data with the pgmm function or any other system Gmm function?
I would really appreciate some help here, was looking for hours for a solution. Thanks a lot!