A lot of my friends, economists or econometricians, do not really like the concept of Granger causality. To some extent, the essence of the concept of Granger causality is expressed in the Latin locution post hoc ergo propter hoc. This means: after this, therefore because of this. In numerous instances, that is a fallacy, a simple temporal precedence. But not always!
In my last paper, in which I propose a novel identification procedure for geopolitical turning points, I use a monthly database of geopolitical relations of China with its 12 main trade partners:
Saadaoui, J. (2025). Geopolitical Turning Points and Oil Price Responses: An IV-LP Approach. Available at SSRN: 5366829.
One question that arises when you look at the data is, are there strategic alliances in setting the diplomatic relations with China?
The answer is yes to put it bluntly. How can I show that? First, I have to look at the data. Some countries have fluctuating relations, like the US and Japan; some have known a recent deterioration, like Australia and the UK, and to a lesser extent Germany and France; others have known a lot of improvements in the 1990s and have known a recent deterioration, like South Korea and India; finally, a group of countries maintain stable and good geopolitical relations with China, namely Indonesia, Vietnam, Pakistan, and Russia.
Strategic military alliances exist in the world; consequently, some bilateral geopolitical relations should be driven by other ones. To accomplish this, I use a VAR(7) model with the change of bilateral relationship to ensure stationarity:
Instead of p-hacking (do not do that), we are going to focus on the chi-squared values that have very large values/very low p-values. The results suggest a strategic alliance between the US, the UK, South Korea, and Australia in setting their geopolitical relations with China. Pretty reasonable from a geopolitical viewpoint. The change in relations of Russia with China precedes those of India and Indonesia, reflecting improvements in the 1990s:
With some good data and geopolitical knowledge, Granger causality may be very relevant.
What happens to the results in the paper controlling for the change of relations with China with other partners than the US? The reaction of oil prices is the same:
You can read the paper if you are interested:
Saadaoui, J. (2025). Geopolitical Turning Points and Oil Price Responses: An IV-LP Approach. Available at SSRN: 5366829.