A multivariate cointegrated modelling approach in testing temporal causality between energy consumption, real income and prices with an application to two Asian LDCs

Unlike previous studies on the casual relationship between energy consumption and economic growth, this paper illustrates how the finding of cointegration(i.e. long-term equilibrium relationship) between these variables, may be used in testing Granger causality. Based on the most recent Johansen's multiple cointegration tests preceded by various unit root or nonstationarity tests, we test for cointegration between total energy consumption, real income and price level of two Asian LDCs: Thailand and Sri Lanka. Nonrejection of cointegration between variables rules out Granger noncausality and implies at least one way of Granger-causality either unidirectional or bidirectional. Secondly, by using a dynamic vector error-correction model, we then analyse the direction of Granger-causation and hence the within-sample Grangerexogeneity or endogeneity of each of the variables. Thirdly, the relative strength of the causality is gauged (through the dynamic variance decomposition technique) by decomposing the total impact of an unanticipated shock to each of the variables beyond the sample period, into proportions attributable to shocks in the other variables including its own, in the multivariate system. Finally, these response paths of shocks to the system are traced out using impulse response graphs. Results based on these four tools of methodology, broadly indicate that all three variables are cointegrated and exhibit two common trends within each system. Energy consumption seems to be relatively exogenous as neither income nor prices seems to Granger cause this variable via any of the channels where potential casuality may occur. Though, energy consumption itself plays an important role in influencing income and prices by varying degrees of significance for each country. Overall, shocks to the system seemed to have had a more sustained if not pronounced effect in Thailand than in Sri Lanka.