Granger Causality test
Posted: Sun Apr 07, 2013 5:34 pm
Hi! I am constructing a model of GDP regressed on 6 independent variables.
After testing for stationarity and running the base regression with stationary variables, I would like to check whether there is causality running from each tax variable onto GDP and vice-versa. But how do you choose the lag length for this test?
Thanks :)
After testing for stationarity and running the base regression with stationary variables, I would like to check whether there is causality running from each tax variable onto GDP and vice-versa. But how do you choose the lag length for this test?
Thanks :)