Energy-income causality in OECD countries revisited: The key role of capital stock

This paper applies a recent advance in panel analysis to estimate the panel cointegration and panel vector error correction models for a set of 22 OECD countries using annual data covering the period 1960-2001. We investigate the relationship between energy consumption and income using an aggregate production function and controlling for the capital stock, as well as by exploring the dynamic directions of the causality among these three variables. We firstly obtain solid and convincing evidence of a fairly strong long-run equilibrium relationship among them. Secondly, it is found that the capital stock is much more productive than energy consumption. Third, it is observed that neglecting the impact of the capital stock on income tends to overestimate the effect of energy consumption. Finally, the panel causality test shows bi-directional causal linkages exist among energy consumption, the capital stock and economic growth. Overall, the findings reveal that the capital stock plays a critical role in realizing the dynamic relationship between energy and income.