What is GHG Accounting? Further Thoughts on Responsibility & Causation

November 20, 2023, by Michael Gillenwater

Installment N.3 bis


This installment supplements the previous installment on emissions allocation. Here I address two conceptual hurdles that you may be struggling to jump over—responsibility and causation. Understanding these concepts is fundamental to achieving a deeper understanding of GHG accounting.

To review, “allocation” refers to the boundary setting rules and norms that assign responsibility for emissions to an accounting subject—such as a company, country, city, facility, or product—when applying physical allocational accounting methods (i.e., preparing emission inventories).[1] In other words, in the context of corporate reporting, we are choosing what emissions a company reports and, therefore, which emissions they have been allocated responsibility.

So then, what do we mean by “responsibility”? Should we interpret responsibility in the sense of placing blame? Emitting pollution is a bad thing and those that are involved in doing bad things should be responsible for what they do. Or, should we interpret responsibility in the sense of establishing accountability? Is there a difference between the two?

The customary character of the “blame”-framing of responsibility is non-exclusive and tied to our beliefs concerning causation (i.e., causal thinking). If you throw a rock at a glass window and it breaks, then we believe that you caused the window to break and are therefore to blame for the window breaking. But what if you and nine other people all simultaneously throw rocks at the window and the window breaks? And what if ten additional people knowingly gathered and provided the rocks for each of you to throw. The blame-framing of responsibility would suggest that all twenty (i.e., ten collectors and ten throwers) of you are to blame because all of you contributed to causing the window to break. This framing underlies the current approach to indirect (Scope 3) emissions under the GHG Protocol. Every company that has some causal influence on emissions anywhere in a value chain is deemed to have responsibility for those emissions, and therefore should report them, even though it results in an undefined large number of companies each reporting the same emissions.

Alternatively, the customary character of the “accountability”-framing of responsibility is exclusive and consciously independent from causal beliefs. If I am managing a project with many tasks, I will assign each task to a specific employee who will “own” that task and be held accountable for its completion. However, it is understood that as a project team, we will help each other and surely contribute to many tasks, including those that are not assigned to us. Responsibility in this framing is structured with exclusivity because accountability means little if everyone is viewed as responsible for every task.[2]

The non-exclusive and causal framing (i.e., blame-framing) of responsibility and allocation, that is applied in the current approach to Scope 3 emissions, is what I mean when referring to the need to eschew consequential thinking in physical allocational GHG accounting (to refresh on these terms revisit installment #1). Instead, GHG inventory accounting should be grounded in the accountability-framing of responsibility. At this point, you may be asking: why should we consciously eschew grounding our GHG inventory thinking in terms of blaming companies for any emissions we interpret them to influence? To answer this question, we will have to descend into a bit of a metaphysical discussion of causation.

We can comprehend causation in two ways: single-factor and multi-factor. With single-factor causation, we view an action as fully and solely causing the release of emissions. No other actions from other subjects have a causal relationship with these emissions. Although conceptually clean and simple, I can promise you that there is not a single unit of GHGs emitted to the atmosphere that we can say had a unitary cause. Everything in the physical universe, such as an indirect emission source in a value chain, is a function of a long and complex chain of causal factors. So, if we use single-factor causation as the basis for our indirect GHG emissions allocation logic, then the GHG inventory for every company would include no indirect emissions because the actions of no company are the SOLE cause of any unit of emissions. We can therefore remove single-factor causation as a feasible basis for allocation.

In reality, every unit of indirect emissions has an array of actions that influenced its release. Some actions may have had a minor influence and some a major influence. With multi-factor causation, each unit of emissions released has many actions influencing it and potentially many subjects associated with each of those actions. Taking this concept to its logical extreme, a company could report any unit of emissions released anywhere in the world at a given time because causal chains of influence can be infinitely extended (see installment #3 for further explanation). Such an approach to assigning responsibility is taken by the GHG Protocol Scope 3 guidance. But, it fails to establish meaningful accountability, as accountability is increasingly diluted as the number of subjects assigned responsibility for an emission source expands.

As an alternative, could we construct indirect emissions allocation rules with exclusivity based on which action is deemed to have the “most” influence on the release of each unit of emissions? Then allocate that unit of emissions to the subject associated with that dominant action. Or maybe we should assign fractions of each unit of emissions based on fractions of causation? Company A’s actions caused 5% of this tonne, Company B’s actions caused 10%, etc.

I will simply assert that both of these ideas are ludicrous unless we want to establish a new accounting field for causality. If we insist on applying causal thinking to answer our emissions allocation questions, then we turn emissions inventory accounting into an enormously impractical and highly subjective causal influence mapping exercise. The upshot of this exposition is that the blame-framing of responsibility (i.e., causal thinking) usefully addresses none of our allocation issues for GHG inventory accounting. In using causal thinking for allocation, we have both made solving the problem of allocation more difficult, AND we have fostered deep confusion between consequential and allocational accounting methods.

Describing a GHG inventory as a causal analysis and a company as a collection of actions is an intellectual dead-end for the purpose of GHG accounting. We can recognize that actions have subjects associated with them, and that physical processes that release emissions are the result of actions. Nonetheless, we should not, and need not, draw on causal thinking to justify allocation choices.

I argue, instead, that allocation rules require there to be a physical connection between the subject and the emission source and be formulated with exclusivity (i.e., allocated to one subject) in proportion to the matter or energy flows of that connection. Critically, the justification for requiring a physical connection should be normative (i.e., an accepted standardized rule). And that this norm not be justified by assumptions that physical connections “cause” specific units of emissions to be released.[3]

Another normative rule should also be that allocation of indirect emissions needs to achieve additivity across a designated population of companies (e.g., within a common industrial sector). Departing from responsibility for a moment, the goal here is to enable comparability in reporting across similar companies within an industry. Beyond these two rules, other GHG accounting rules should be selected based on the needs of the intended application of the reported physical inventory estimates. This latter point means that our accounting rules should be designed for a specific type of application, to accomplish a specific goal, and that we should cease believing that one corporate accounting protocol or standard can serve all functions.

You may now be thinking that causation must still matter for GHG accounting. And it does. But physical allocational GHG accounting (i.e., GHG emission inventorying) is the wrong tool to address causation. Luckily, we have another set of GHG accounting methods that are designed to address causation: consequential or intervention methods. What we are missing, though, is a well-established reporting framework for companies to estimate and disclose the avoided emission (and enhanced removal) impacts of their interventions. To make space for such a new intervention impact reporting framework, we need to intellectually distinguish it (and causal thinking more broadly), from GHG inventory reporting. Building this kind of new intervention impact reporting framework for companies should be an additional focus of our community’s work to update GHG reporting protocols over the coming year.

[1] Note that I use the term “allocational” in place of “attributional”, as the latter term is problematic and should be replaced with the former. See here for a discussion of why. I will also use the term “emissions” for brevity, but you should interpret this to mean both GHG emissions and removals.

[2] While behavioral scientists or linguists might dive deeper into the perceived differences between blame and accountability; using this simple characterization helps make explicit our implicit thinking about GHG accounting. Again, blame-framing is non-exclusive and causal, while accountability-framing is exclusive and not principally causal.

[3] One could imagine an alternative norm for allocation rules that requires only a financial connection or both a financial and physical connection. The outcome of relying purely on a financial connection as the norm would be an approach that allows a market for emission attributes (i.e., a market for emission factors in which companies can “purchase” the inventory results of their choice regardless of their physical activities). The outcome of a norm based on both financial and physical connections is an approach that fails to address processes that clearly should be included in inventory accounting boundaries, as their physical connection is apparent, but have no financial recording associated with the subject.

Cover photo from Nosha on Flickr (CC BY-SA 2.0 DEED), cropped image.

Leave a Reply

Your email address will not be published. Required fields are marked *