An investigation of issues relating to where energy should enter the production function

This paper examines the impact of imposing different separability assumptions in the specifications of the standard hierarchical KLEM production function in a computable general equilibrium (CGE) model. The appropriate means of introducing energy to sectoral production functions in CGE models has been a source of debate for a number of years. However, while modellers often subject their model results to sensitivy analysis with respect to the values associated with elasticities of substitution between inputs, it is rarely the case that the structure of the production function is subjected to testing. However, the chosen structure reflects the modeller's view about elasticity between different inputs and will have implications for model results wherever there are changes in relative prices. We illustrate our argument by introducing a simple demand shock to a CGE model of the Scottish economy (targetted at the energy supply sector) under different assumptions regarding the structure of the KLEM production function and separability assumptions therein. Specifically, we conduct both systematic and random parameter variation within alternative KLEM production structures, examining the impacts on a number of model outputs, though with primary focus on energy use in production. We find that if energy is introduced to the value-added rather than intermediates nest there is greater variation in energy use in production in response to the demand disturbance.