Digital investment and industrial energy efficiency in Central and Eastern Europe: A stochastic energy demand frontier analysis

This study investigates the role of digital technologies in shaping industrial energy efficiency in post-accession Central and Eastern European (CEE) countries, distinguishing ICT-equipment investment from R&D/software investment. Using panel data for 15 industrial sectors across 9 CEE economies over the period 2012–2020, the study estimates a stochastic energy demand frontier model that is implemented within a consistent true fixed effects framework. Estimated efficiency scores indicate a relatively high average energy efficiency (0.89), suggesting potential energy savings of approximately 11%. Energy demand responds positively to output and negatively to energy prices, while heating-degree days exert a substantially larger and more robust effect than cooling degree days. ICT-equipment intensity is robustly associated with higher energy efficiency, while R&D and software intensity is robustly associated with lower energy efficiency, using sector-level investment data from EUKLEMS capital accounts; both the frontier and inefficiency determinants vary systematically with sectoral energy intensity and EU ETS exposure. Findings suggest that hardware-based digital capital deepening supports energy efficiency, whereas the adjustment costs associated with R&D-driven innovation may weigh on measured efficiency, even if only temporarily. The results underscore the importance of distinguishing digital investment types in the design of energy and industrial policy in post-transition European economies.