The responsiveness of energy demand to economic factors: econometric analyses on the sectoral level of OECD countries

Bernstein, Ronald; Madlener, Reinhard (Thesis advisor); Balleer, Almut (Thesis advisor)

Aachen : RWTH Aachen University (2023)
Dissertation / PhD Thesis

Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2023


The thesis aims at gaining insights regarding the linkages between energy demand and other economic factors in order to provide scientific guidance for energy planning and policy designs. Especially in the light of the current debate and concern regarding anthropological causes of global warming and its effects on the environment, these topics have risen again on the agenda of policy-makers and researchers alike. The central research questions which this thesis aims at shedding light on can be summarized as follows: [1] What is the response of energy demand (differentiated by single energy carriers) to changes in energy price and economic activity in miscellaneous sectors of developed (OECD) countries? [2] Does causality run from energy consumption to economic activity or vice versa? Or is the relationship characterized by a feedback system? [3] How do information and communication technologies (ICTs) impact energy consumption in production processes (at the sectoral level)? These research questions are tackled empirically by applying various techniques within a top-down econometric approach. In this regard, the focus lies on the cointegration methodology. By explicitly taking into account stochastic trends inherent in aggregate time series (and panel) data on energy consumption, prices and other variables, long-run equilibrium demand relationships are estimated. By measuring the response of energy demand to changes in energy price, economic activity and other factors, this methodology allows to draw conclusions with regard to the long-run adjustment processes modelled within an error-correction mechanism framework. Moreover, Granger-causality tests further provide an indication for the direction of causality between the variables considered. The first three papers of the thesis investigate the elasticity of electricity demand with regard to energy price and economic activity in the German manufacturing industry, and the OECD residential sector, and the natural gas demand elasticity in the residential sector of OECD countries by applying the maximum likelihood system approach, the fully modified OLS and dynamic OLS group-means panel estimation framework, and the auto-regressive distributed bounds testing procedure, respectively. The first two papers additionally perform tests on Granger-causality between the variables considered. The fourth paper differs from the other papers in that it applies a fixed effects estimator that is based on a factor demand model to estimate the impact of disaggregated ICT capital on energy demand. With regard to the research questions [1] and [2] above, some consistent findings and general conclusions can be drawn from the work concerned with estimating energy demand elasticities with regard to economic activity and price:•The long-run energy demand elasticities with regard to economic activity tend to be near unity, with mean values of 1.14 for industrial electricity demand, 0.96 for residential electricity demand, and 0.94 for residential natural gas demand. • The long-run energy demand elasticities with regard to energy price are in most cases negative and rather small (or statistically not significant from zero), with mean values of -0.22 for industrial electricity demand, -0.39 for residential electricity demand, and -0.51 for residential natural gas demand. • The short-run energy demand elasticities, both with regard to economic activity and energy price, generally have the same sign as their long-run counterparts, but are smaller in magnitude. • Tests on Granger causality find evidence for causality running from economic activity to energy consumption in both the industrial as well as the residential sector. Moreover, some evidence is also found for an opposite causal relationship, i.e. from energy use to economic activity in both sectors. Altogether, regardless of the sector or energy type considered, the results imply that the steering effect of tax-induced price increases on energy demand has a very limited potential for energy conservation, and hence a reduction of GHG emissions. Furthermore, the findings suggest that reductions in electricity consumption are associated with a trade-off concerning economic growth in the residential sector, as well as some subsectors of the manufacturing industry. With regard to research question [3] introduced above, the findings obtained are twofold. On the one hand, the analysis provides evidence for an electricity-saving effect on production induced by communication technologies in all of the sectors considered. On the other hand, the effect of computers and software on the electricity intensity of industrial production is not that clear-cut, but rather seems to be strongly dependent on the sector-specific production processes involved. Overall, the net effect of ICT diffusion on electricity intensity of production appears to be in favor of an enhancement of electricity efficiency in production.