We Need Better Carbon Accounting—Here’s How To Get There
Any effective system of greenhouse-gas accounting needs to measure each company’s supply-chain carbon impacts accurately, providing visibility and incentives for it to make more climate-friendly product-specification and purchasing decisions.
The Rocky Mountain Institute reports that the average company’s supply-chain greenhouse-gas (GHG) emissions are 5.5 times higher than the direct emissions from its own assets and operations. Any effective system of GHG accounting, therefore, needs to measure accurately each company’s supply-chain carbon impacts, providing visibility and incentives for it to make more climate-friendly product-specification and purchasing decisions.
A recent Harvard Business Review article, “Accounting for Climate Change,” noted how the current dominant system for carbon accounting, the GHG Protocol, misses this critical point by allowing companies to guestimate upstream and downstream emissions. To address this shortcoming, it introduced an E-liability accounting system, based on well-established practices from inventory and cost accounting, for accurately measuring GHG emissions across corporate supply chains.
Since the article’s publication, the authors have had dozens of conversations with corporate executives, consultants, regulators, and standard-setters about the E-liability system. Many have expressed frustration that something like it has not been introduced sooner. This follow-up piece describes the basic flaw inherent in the GHG Protocol, explains why it has persisted, and offers a way forward for robust carbon accounting that does not involve rescinding the protocol, which has been widely embedded in many global climate agreements. It concludes by identifying which companies stand to gain most from accurate GHG accounting and could be early adopters of the E-liability system.
The Current GHG-Accounting Standard Discourages Supply-Chain Decarbonization
Introduced in 2001, the GHG Protocol has become the de facto global accounting standard for measuring an entity’s direct, upstream, and downstream GHG emissions. The Protocol’s Scope 1 standard provides the basis for valid accounting of an entity’s direct GHG emissions, an important feature because these are the only corporate emissions that actually go into the atmosphere.
But the Protocol falls short in its Scope 3 standard, which requires a company to estimate the Scope 1 emissions of all its direct and indirect suppliers and customers. Even apart from the multiple counting of the same emissions inherent to this approach, the previous article expressed skepticism about the standard’s feasibility. Most companies know only a few of their non-Tier-1 suppliers and customers well enough to get meaningful data from them. Yet the protocol expects companies with diverse product lines to gather emissions data from all their multiple-tier customers and suppliers for each line—a fiendishly complex task.
The near-impossibility of measuring Scope 3 emissions forced the protocol standard-setters to allow companies the option to use industry and regional averages, rather than measure the specific emissions produced by their actual suppliers, distributors, and customers. Although the protocol expresses a preference for “primary data,” defined as “provided by suppliers or other value chain partners related to specific activities in the reporting company’s value chain,” it allows the use of secondary data “in some cases, [when] primary data may not be available or may not be of sufficient quality.” Secondary data is defined as “industry-average data (e.g., from published databases, government statistics, literature studies, and industry associations), financial data, proxy data, and other generic data.”
Unsurprisingly, “some cases” has, in practice, become “for all cases.” But allowing companies to use average rather than specific and traceable data fundamentally undermines the integrity of Scope 3 measurements. Imagine a financial accounting standard that allows a company to use industry-average raw-material costs rather than actual invoiced raw-material costs. Would such a financial report, based on average rather than actual profit margins, be acceptable to shareholders, financial analysts, and tax authorities? Yet this is the standard set by the protocol for reporting Scope 3 emissions.
Fortunately, Scope 3 reporting is currently voluntary, and most companies skip reporting on their supplier and customer emissions. Even among those that do, there is skepticism; IBM, for example, notes “the assumptions that must be made to estimate Scope 3 emissions in most categories do not enable credible, factual numbers.” A few companies, such as Mars, voluntarily use industry-average data to comply largely with the Scope 3 standard. Most companies reporting on their Scope 3 emissions, however, are quite selective, cynics would say opportunistic, about what they report. Google and Microsoft, although in the same industry and generally considered leaders in climate accounting, report different categories of indirect GHG sources.
US regulators also seem to have reservations about the validity of Scope 3 reporting. Last month, the US Securities and Exchange Commission proposed that certain registrant companies must provide audited Scope 1 and 2 climate disclosures by 2024. But it also provided a litigation safe harbor to any companies that voluntarily provide Scope 3 disclosures, an implicit acknowledgement that such disclosures would be unreliable and unauditable.