This study analyzes energy intensity trends and drivers in 40 major economies using the WIOD database, a novel harmonized and consistent dataset of input-output table time series accompanied by environmental satellite data. We use logarithmic mean Divisia index decomposition to(1) attribute efficiency changes to either changes in technology or changes in the structure of the economy, (2) study trends in global energy intensity between 1995 and 2007, and (3) highlight sectoral and regional differences. For the country analysis we apply the traditional two factor index decomposition approach, while for the global analysis we use a three factor decomposition which includes the consideration of regional structural changes in the global economy. We first show that heterogeneity within each sector across countries is high. These general trends within the sectors are dominated by large economies, first and foremost the United States. In most cases, heterogeneity is lower within each country across the different sectors. Regarding changes of energy intensity at the country level, improvements between 1995 and 2007 are largely attributable to technological change while structural change is less important in most countries. Notable exceptions are Japan, the United States, Australia, Taiwan, Mexico and Brazil where a change in the industry mix was the main driver behind the observed energy intensity reduction. At the global level we find that despite a shift of the global economy to more energy-intensive countries, aggregate energy efficiency improved mostly due to technological change.
Voigt, Sebastian, Enrica De Cian, Michael Schymura and Elena Verdolini (2014), Energy Intensity Developments in 40 Major Economies: Structural Change or Technology Improvement?, Energy Economics 41 , 47-62