Climate Protection Potentials of Digitalized Production Processes: Microeconometric Evidence?

What Drives the Relationship Between Digitalization and Industrial Energy Demand? Exploring Firm-Level Heterogeneity

“…To the best of our knowledge, no econometric study to date examines absolute energy use in the manufacturing sector, yet empirical evidence exists with respect to changes in energy intensity. Applying a large administrative panel data set on manufacturing firms, Axenbeck and Niebel (2021) observe only marginal average energy intensity improvements related to software usage. Besides, the authors find that, even though overall effects are small, the relationship is more pronounced in energy-intensive industries, which indicates effect heterogeneity with respect to different production processes.…”

“…8 Not without reason, it is a commonly used indicator at the firm level (cf. Almeida et al, 2020;Bessen and Righi, 2020;Axenbeck and Niebel, 2021;Barth et al, 2022). We also integrate a tangible capital stock K. We calculate both capital stocks by applying the perpetual inventory method (PIM), which allows generating a productivity-relevant capital stock (c.f.…”

“…BMWK former BMWi (heating oil, retrieved on: 01.04.2020). For a more detailed description on sources for energy prices see Axenbeck and Niebel (2021).…”

“…Previous empirical studies on the relationship between ICT and environmental impacts mainly attempt to prove a linear and directional link between ICT and CO 2 emissions (Zhang and Liu, 2015;Chen et al, 2019;Kopp and Lange, 2019) or energy efficiency and energy use (Collard et al, 2005;Bernstein and Madlener, 2010;Schulte et al, 2016;Axenbeck and Niebel, 2021;Huang et al, 2022). However, as the relationship may be heterogeneous or non-linear (Ben Lahouel et al, 2021;Taneja and Mandys, 2022;Xu et al, 2022), standard regression models fall short of fully uncovering the complexity of the relationship.…”

“…Note here that electricity consumption and fossil fuel use do not sum up to energy use, since non-electric non-fossil energy, such as biomass, cannot be accounted to either of the two.8 For a detailed description of the capital stock approximation as well as for a descriptive analysis of the software capital indicator, we refer toAxenbeck and Niebel (2021).…”