Old wine in new bottles: Testing causality between savings and growth

The paper uses the methodology of Toda and Yamamoto to test for causality between growth and savings in order to avoid problems and possible misleading inferences associated with the asymptotic nature of Granger causality testing in time-series studies. The relationship between gross domestic product, gross domestic savings and private savings was examined for India and Sri Lanka. We found no causality between GDP growth and private savings in India, while it appears that there is bidirectional causality between private savings and growth in Sri Lanka. We conclude that existing 'evidence' on the subject should be treated with caution, given the inappropriateness of the econometric methodology adopted in most of the previous empirical studies using time-series data.