Niklas Letmathe

2026

Carbon accounting has become central to corporate reporting, regulation, and mitigation, but the allocation of greenhouse gas emissions to products, organizations, supply-chain actors, and territories remains error-prone. This thesis examines the barriers and errors that affect carbon-emission allocation across these contexts and assesses whether concepts and methods from Cost Accounting can help reduce them. The analysis combines a systematic literature review supplemented by six targeted studies, resulting in an analytical sample of 47 Carbon Accounting studies, with a theory-guided examination of Cost Accounting literature anchored in the error framework of Datar and Gupta (1994). The review identifies comparability and standardization, data availability, and system boundaries as the most frequent barriers. The identified errors fall into four categories: measurement, specification, aggregation, and errors beyond the Datar-and-Gupta framework. The first three categories show substantive parallels to Cost Accounting, where determinant-based system design, cost-driver selection, Activity-Based Costing, and cost-pool design offer potentially transferable approaches. Their transfer is bounded by spatial heterogeneity, the choice between process-based and input-output accounting, and the trade-off between detail and implementation effort. Errors in the fourth category, dominated by double counting, are interpreted as a multi-entity attribution problem that lies outside the classical framework because the reviewed Cost Accounting mechanisms rest on an intra-firm authority that value chains do not provide. Transactional Carbon Accounting architectures extend allocation beyond the firm, but their practical implementation depends on standardization, verification, and incentives. Cost Accounting therefore provides a substantive reference for the design of carbon-allocation rules, without replacing the institutional arrangements on which effective Carbon Accounting depends.