Systematic bias in Scope 3 emissions accounting: evidence from firm-level supply networks

Scope 3 emissions often constitute the largest share of corporate greenhouse gas footprints, yet their accounting remains highly sensitive to incomplete supplier-level information and the replacement of aggregated emission-intensity proxies. Here, using Toyota Motor Corporation as a firm-level case study, we examine the data gap of third-party supply-chain databases and assess how the granularity and source of emission-intensity data shape upstream Scope 3 estimates. We construct a five-tier supply-chain network for this company, drawing upon publicly available supplier-customer relationships, firm financial reports, geographic information and a gravity model used for trade volumes prediction, which is therefore conservative, transparent, and reproducible for shareholders to implement and verify. Benchmark with the network constant, four emission-intensity scenarios, ranging from firm-level intensities to subindustry-, sector-, and MRIO-based sector averages are compared. Results show that sector-average emission intensities would overestimate the Scope 3 emissions; the magnitude of the overestimates hinges on the data scales and database selections. This study further reveals a pronounced emissions concentration, with the first-tier suppliers accounting for around 60% of upstream embodied emissions across the five supply tiers. A similar emissions concentration also exists among a small number of carbon-intensive suppliers within each supply tier; most of them are engaged in raw materials provision and energy-intensive activities. Firm-level analysis of emission hotspots also demonstrates that trade network reconstruction can improve the interpretability and decision relevance of corporate Scope 3 accounting by revealing where embodied emissions are concentrated and how mitigation efforts can be prioritized across global supply networks.