Overcoming the shortcomings of energy intensity index: a directional technology distance function approach

In multilateral comparisons of environmental performance over time, energy intensity measures, especially "real" energy intensity computed either by index decomposition approach or structural decomposition approach, are the most commonly used measures. Recently, researchers also resort to production-theoretical approach, which relies on data envelopment analysis techniques, to decompose energy intensity changes over time into their subcomponents. While their intuitiveness and computational ease make these indices attractive, their time series properties create considerable challenges in performing informative and fair comparisons among the energy efficiency levels of units considered. Furthermore, the resultant measure of energy intensity in these studies is still the inverse of a partial factor productivity (PFP) measure, i.e., energy productivity, that does not take into consideration compositional differences between inputs of the units being compared (which are also subject to change over time) and that ignores the type of substitution among inputs and, hence, makes it a measure that disguises rather than illuminates. The theoretical part of this paper shows how one can overcome the shortcomings of the energy intensity measure by constructing a new energy index using directional technology distance functions. The new index constructed in this study not only overcomes the shortcomings of the energy intensity measures but also satisfies the axiomatic properties of index numbers that are laid down by Fisher. An empirical application on U.S state-level agricultural sectors further complements existing studies.