Cointegration and Causal Analysis between Energy and Output: Evidence of China

This paper uses a Chinese aggregate production technology which includes energy, capital and labor as inputs of production and examines their relationships in terms of cointegration and causality. Conventional cointegration and causality approaches are invalid for variables with a mixture of order of integration. Hence the autoregressive distributed lag (ARDL) bounds test approach identifies the existence of cointegration among the variables. The Toda-Yamamoto approach is conducted to examine direction of Granger causality between them. The empirical findings indicate a long-run equiliburium of output, labor, capital and energy, and energy is an important variable in the production function such as labor and capital. We found bi-directional Granger causality between energy and output, and there is Granger causality running from labor to output. The result supports that energy is an important variable in the aggregate production, and energy is closely related to output in terms of Granger causality.