PRACTICAL ASPECTS OF CAUSAL INFERENCE - AN INDIAN APPLICATION

This paper is guided by two objectives-to present a brief review of major recent developments in the theory of Granger-causality and to highlight the critical role of three important empirical issues viz. seasonality, lag-selection and serial correlation in any causality-testing exercise. The relationship between money and prices in India serves as the focal point for the discussion of these issues. It is found that causal inferences are sensitive to the deseasonalisation and lag-selection procedures adopted, as well as the causality test chosen. In the Indian context, the only conclusion which emerges in a robust fashion, is that of "narrow" money causing prices, lending a measure of support to monetarism.