One of the most consequential (no pun intended) methodological choices in a Life Cycle Assessment is not the functional unit or the impact category, it is whether the study is modelled attributionally or consequentially. The two approaches can be built from the same inventory data and still arrive at meaningfully different results, because they are answering different questions.
What is Attributional LCA (ALCA)?
Attributional LCA describes a product system as it exists today, or as it existed on average over a historical period. It asks: what share of the world’s current environmental burden can be attributed to this product? It uses average, supplier mix data, and where a process produces more than one output, allocates the shared burden between them using rules such as mass, economic value, or energy content, following the allocation hierarchy set out in ISO 14044.
Because it describes an existing, static system, ALCA is well suited to reporting: Product Carbon Footprints, Environmental Product Declarations (EPDs), and most corporate sustainability disclosures are built on attributional models. It answers the question “what is our current footprint?”
What is Consequential LCA (CLCA)?
Consequential LCA instead asks: what actually changes in the wider world if we make this decision? Rather than allocating an average share of existing burdens, it models the marginal, or “last unit”, supplier or technology that will actually respond to a change in demand, and traces the knock on effects this triggers elsewhere in the market, including effects on co-products, substituted products, and constrained supply.
This is the modelling approach most closely associated with the research led by Bo Weidema, and colleagues including Jannick Schmidt, at what is now the 2.0 LCA consultants group and Aalborg University’s sustainability research, who have argued that attributional models can obscure the real physical and market causalities that a decision sets in motion, since average data reflects the past rather than the actual consequence of a choice made today.
CLCA is the more appropriate tool when the question is genuinely about the future: should we switch suppliers, redesign a product, or change a policy, and what would actually happen as a result? It answers the question “what happens if we change something?”
| Attributional LCA | Consequential LCA |
|---|---|
| Describes the system as it exists today | Describes how the system will change in response to a decision |
| Uses average, supplier mix data | Uses marginal, “last unit” supplier data |
| Allocates shared burdens between co-products | Expands the system boundary to avoid allocation |
| Well suited to reporting and disclosure (PCFs, EPDs) | Well suited to redesign, sourcing, and policy decisions |
| Answers “what is our footprint today?” | Answers “what happens if we change this?” |
| Simpler to communicate and audit | Requires identifying affected markets and marginal suppliers, which carries more modelling uncertainty |
Why the Debate Is Still Open
Which approach is “correct” remains a genuinely active and unresolved debate within the life cycle assessment community, discussed extensively in journals such as the International Journal of Life Cycle Assessment and by researchers across the field, not a settled question with one right answer. Attributional practitioners point to the practicality, comparability, and auditability of average data. Consequential researchers, including Weidema and colleagues, argue that only a consequential model reflects the actual physical and economic causality triggered by a decision, and that using attributional background data inside a decision oriented study can introduce meaningful errors.
In practice, most industry LCAs default to attributional modelling, largely because it is what current international standards, databases, and reporting frameworks are built around, while consequential modelling remains more common in academic research and in studies specifically designed to inform strategic decisions or policy.
Our Position
We don’t treat this as a question with a single right answer, because the LCA community itself hasn’t settled it. Our recommendation depends on what the study needs to do:
- If you need a Product Carbon Footprint, EPD, or corporate disclosure, industry practice, comparability, and current standards point strongly toward an attributional model.
- If you are trying to genuinely improve a product, evaluate a redesign or supplier switch, model how market mechanisms and future scenarios would respond to a decision, or are doing a comparative or prospective LCA, we recommend consequential modelling, since it is built specifically to answer that kind of question.
Our own practice leans toward consequential modelling wherever the study is meant to inform a decision rather than simply report a status quo.
We’re equally comfortable working in either paradigm, and we’ll always be transparent with you about which one a given study uses, and why, so the result is interpreted correctly.