Hi,
I have data on foreign investment in different countries at the quarterly level (n=60, t=54).
I wish to estimate a dynamic panel model of the determinants of investment using a FE model.
For some reason, Stata drops the time dummy of every first quarter of each year due to collinearity (2001Q1, 2002Q1, 2003Q1, and so forth).
It doesn't drop these time dummies when I do not use the lagged dependent variable, but this doesn't help much as I need to use the lagged dependent variable.
Any idea how to fix this?
I have data on foreign investment in different countries at the quarterly level (n=60, t=54).
I wish to estimate a dynamic panel model of the determinants of investment using a FE model.
For some reason, Stata drops the time dummy of every first quarter of each year due to collinearity (2001Q1, 2002Q1, 2003Q1, and so forth).
It doesn't drop these time dummies when I do not use the lagged dependent variable, but this doesn't help much as I need to use the lagged dependent variable.
Any idea how to fix this?
