Non-parametric frontier approach to modelling the relationships among population, GDP, energy consumption and CO2 emissions

In this paper, a non-parametric approach based in Data Envelopment Analysis (DEA) is proposed as an alternative to the Kaya identity (a.k.a ImPACT). This Frontier Method identifies and extends existing best practices. Population and GDP are considered as input and output, respectively. Both primary energy consumption and Greenhouse Gas (GHG) emissions are considered as undesirable outputs. Several Linear Programming models are formulated with different aims, namely: a) determine efficiency levels, b) estimate maximum GDP compatible with given levels of population, energy intensity and carbonization intensity, and c) estimate the minimum level of GHG emissions compatible with given levels of population, GDP, energy intensity or carbonization index. The United States of America case is used as illustration of the proposed approach.