Corporate carbon footprinting is all the rage, scope 1, 2 and 3, which means taking the supply chain and potentially also the customer into account. Looking at these footprints in aggregate, it becomes clear that double-counting becomes a massive problem. Its solution is explored in a new paper.
Prominent scientists have therefore argued that carbon footprints should only be calculated for consumer goods, not for anything sold to other companies.
The argument that the footprint of an intermediate product is not defined is not correct, as I have shown before, because the derivation of the footprint via the Leontief inverse assumes and hence requires that intermediate inputs have the same footprint (or economic multiplier) as products sold to final consumers.
However, there are good questions to look at transactions within the value chain and ask, what was required to produce this? Of course, there is double counting. If you look at the value chain of steel, you will find mines, train cars, and furnace halls that contain much steel. Hence, some of the product goes to producing itself, or inputs to its own production. If you want to know what are the total emissions from steel production, you need to ignore the inputs of steel to steel production, or subtract the steel that is used anywhere in the value chain. This problem can be solved my clever math.
Economists have solved this problem a long time ago, employing concepts of hypothetical extraction and total flow. Industrial ecologists have independently developed an approach. How do these differ?
In a new paper published in the Journal of Industrial Ecology, Maximilian Koslowski, Kajwan Rasul and I show that these approaches are mathematically identical and fit into a broader framework.
The paper also provides a good mathematical introduction to input-output analysis, covering both the Leontief price and quantity models and based on the system definition of the input-output economy developed by my colleagues Daniel Müller and Stefan Pauliuk.
Further, our JIE paper unites approaches of literature from Ecological Economics, Industrial Ecology, and Structural Economics.
The hypothetical extraction approach and its variants has a lot of potential, especially when combining data from different data sources, or for identifying elements further up in the supply chain that relate to a specific branch. I recommend anybody who works with MRIO analysis to become familiar with it!
I have a rather long presentation of this on YouTube, given as a webinar of the International Input-Output Association in 2023. An updated version of the slides, based on my Earth-Day lecture at Leiden University, is posted on Slideshare.
