Shows how the third-series effect is eliminated with least causal distortion, introducing partial measures of the one-way effect, reciprocity, and association

Illustrates the proposed causal characterization by means of empirical applications to real data sets of the US macroeconomy and Japan's financial economy