I am interested in determining whether electricity consumption causes GDP or vice versa, following the model of Mozumder and Marathe (2007 - https://www.sciencedirect.com/scienc...01421505003320).

I have figured out how to do an augmented Dickey-Fuller and rejected non-stationary among first-differences. I then used 'vecrank' to reject the hypothesis that there is at most 1 cointegrating equation.

Now, however, I am confused as to how to use a VEC model and test for Granger causality. I tried just doing
Code:
vec PCGDP_DIF1 PCEC_DIF1
, but I'm not really sure how to interpret the results.


How should I be proceeding here?