Within-industry Heterogeneity of Emission Productivity: Evidence from Portuguese Manufacturing Firms
Nitschke, J. (2026)
Abstract
Industrial decarbonisation policies are predominantly organised around sectors, targeting energy- and carbon-intensive industries as the main units of regulation. In contrast, a large body of economic literature documents substantial heterogeneity among firms operating within narrowly defined industries.
To examine whether similar heterogeneities can be observed regarding carbon emission performance, this article employs a novel dataset of constructed energy-related Scope 1 and Scope 2 emissions for Portuguese manufacturing firms. The dataset combines administrative financial information with detailed energy-use data, providing unusually broad coverage of firms within narrowly defined industries.
The article analyses how much energy-related emission productivity varies across firms and how much of this variation can be attributed to firm-level versus sector-level factors. Using within-industry dispersion measures, the results show that energy-related emission productivity varies widely across firms operating in the same sector, with substantially larger dispersion than observed for labour, capital, or material productivity.
To investigate the sources of this heterogeneity, a variance decomposition analysis shows that while sector-level factors matter more for energy-related emission productivity than for other productivity measures, firm-level effects remain the dominant source of variation. These findings highlight the importance of further research on within-industry heterogeneity in emission performance and raise questions about the adequacy of climate policies that primarily target sectors rather than firms.
Keywords
emissions, productivity, firm heterogeneity, decarbonisation, manufacturing
Citation
Nitschke, J. (2026). Within-industry Heterogeneity of Emission Productivity: Evidence from Portuguese Manufacturing Firms. Journal of Cleaner Production, 569, 148583. DOI: 10.1016/j.jclepro.2026.148583.
Links
- DOI: https://doi.org/10.1016/j.jclepro.2026.148583
- Free access (available until 29 June 2026): https://authors.elsevier.com/a/1nExU3QCo9uPHm