Can accounting fix global climate disclosure challenges?
As mandatory climate-related reporting gains steam, Harvard Business School’s Robert Kaplan says E-liability accounting addresses key shortcomings that block progress
Environmental, social and governance (ESG) factors are growing in importance but remain difficult to measure and thus to manage. But as mandatory disclosure requirements gain traction globally, a solution to this problem may be emerging from fundamental accounting science and principles.
Professor Robert S. Kaplan, Senior Fellow and Marvin Bower Professor of Leadership Development, Emeritus at Harvard Business School, is co-architect of the E-liability system of accounting for greenhouse gas (GHG) emissions. The system tethers emissions to products and the processes that went into creating them. Each supply chain participant’s emitting liabilities are assigned to its output products and passed downstream to the next customer in the chain, down to the end-use consumer.
The prevailing system of Scope 1, 2 and 3 emissions in the standard-setting GHG Protocol, in contrast, extends a company’s liability to its downstream customers’ emissions – which are virtually impossible for the supplying company to accurately measure or influence. This dilutes the accuracy and accountability for global emissions, according to Prof. Kaplan, who recently delivered the 20th annual Bill Birkett Memorial Lecture, hosted by the UNSW School of Accounting, Auditing and Taxation.
Photo gallery: The 20th annual Bill Birkett Memorial Lecture
The 20th annual Bill Birkett Memorial Lecture. Photos: Stanley Images
Prof. Kaplan and his colleague, Karthik Ramanna, Professor of Business & Public Policy at the University of Oxford’s Blavatnik School of Government, have also proposed accounting-based solutions to key challenges in managing e-assets, such as carbon credits, and in disclosing downstream emissions.
“We can’t manage and improve what we don’t measure, but the science of measurement is not well done by majority vote or committee,” said Prof. Kaplan in his opening remarks. He agreed with the observations of moderator Elysse Morgan of Stafford Capital Partners, who said these well-known axioms are particularly relevant for rethinking the world’s wicked problems, including global approaches to climate-related disclosures.
He also noted that this year’s Birkett lecture was well-timed given the Australian government’s September announcement that it will require climate-related disclosure for large companies beginning 1 January 2025. While such action is a key step in the global objective of reaching net-zero emissions, Prof. Kaplan observed shortcomings in these approaches, unless disclosure is based on fundamental accounting science and principles.
“The big challenge in measuring greenhouse gas emissions is not in measuring the emissions that 95% of companies and large nonprofits, such as hospitals and universities, emit with their own assets,” he explained. “Unless the organisation is in the electricity or materials business, the emissions created in the supply chain for the products and services they purchase are, on average, five-and-a-half times their own emissions. So, the action is not in the so-called Scope 1 emissions of most organisations – it’s in reducing the Scope 1 emissions produced by their suppliers when producing and delivering the products the organisations purchase.”
Rethinking the GHG Protocol
The GHG protocol recognised the importance of supplier emissions when it created its three-scope system more than two decades ago. It recognised that a company’s liability could extend beyond its own emitting assets to include those produced by its upstream energy companies and those in the products and services acquired from its supply chain. But the GHG Protocol also required measurement of emissions from all a company’s downstream customers. Prof. Kaplan stated that he and Prof. Ramanna identified this as a “serious limitation” in the protocol’s framework and application.
To illustrate this problem, he shared a graphic from the GHG Protocol showing all three scopes of emissions entering the atmosphere – Scope 1 direct emissions, Scope 2 from electricity providers and Scope 3 emissions from upstream suppliers (other than electricity) and downstream customers. “That’s curious, because in the history of human endeavour and civilisation, not one molecule of Scope 2 or Scope 3 emissions has ever gone into the atmosphere,” he said. In other words, the emissions created from Scope 1 entities at the source – in the extraction of resources, agriculture, generation of energy, and transportation – are the only ones that enter the atmosphere; the rest is, essentially, bad accounting.
“The challenge, then, is how we accumulate and assign accountability for Scope 1 emissions and not have them counted multiple times by multiple entities,” Prof. Kaplan explained.
Read more: ESG standards are changing. What’s the impact on business?
The GHG Protocol’s limitations stem at least partly from its roots as a voluntary system, which left large downstream companies with the burden of estimating all the emissions contributed by their suppliers. Many of these companies have upwards of 15,000 tier-one suppliers, each of which in turn has multitudes of its own, “going all the way back to the extraction of minerals and hydrocarbons”, Prof. Kaplan said. “Toyota, one of the greatest logistics companies in the world, doesn’t even know the names of 85% of its suppliers because they’re so far upstream. How can it possibly know the emissions those companies create when producing the components that eventually end up in a Toyota vehicle?”
Emissions disclosure’s ‘Copernican moment’
When Profs. Kaplan and Ramanna first looked at the problem three and a half years ago, they immediately saw the impossibility of a downstream company accurately measuring its suppliers’ emissions. While the GHG Protocol accommodates this problem by allowing companies to use industry average emissions factors, Prof. Kaplan said no accounting system “should accept industry-average as a substitute for company-specific data”.
He viewed the E-liability solution as a “Copernican moment.” Standing in a different place and looking in a different direction provided a simple solution to what appeared to be a complex and wicked problem. “Working from fundamental accounting principles, we start at the beginning, with the original producers of Scope 1 emissions – the mining companies producing iron ore and metallurgical coal or the hydrocarbon companies pulling natural gas out of the seabed – that assign their Scope 1 direct emissions to their outputs,” he said. This creates a record of, say, carbon dioxide (CO2) emissions per tonne of metallurgical coal.
Next, when the coal is shipped, the e-liability system allocates the emissions from shipping and transport to all the product on the ship’s deck, and the downstream purchaser, a steel company, will receive a report on the accumulated emissions up to that stage for every tonne of coal it has acquired. The steel company then adds its own Scope 1 emissions from using the coal to produce coils of sheet steel.
The coils can be processed further, with each entity adding its own Scope 1 emissions to those previously embedded, eventually reaching a tier-one supplier, such as a car door manufacturer. This company adds its own (Scope 1) emissions to those embedded from all its suppliers and assigns the total to its outputs, which it then reports that data to its customer, the original equipment manufacturer (OEM) of the car. Ultimately, the end-use consumer will receive an accurate record of all Scope 1 emissions directly created in the production and delivery of the car they purchase.
“This system, when universally applied, provides accurate, auditable data on the emissions for every product and service produced in the global economy,” Prof. Kaplan said.
‘How global decarbonisation occurs’
For companies instituting the e-liability accounting system, the work required to assign their own emissions and the input emissions they acquire to outputs is “very akin” to activity-based costing, a cost-accounting concept Prof. Kaplan helped introduce 40 years ago. Under this approach, the company allocates the emissions in the inputs it purchases and those produced by its production processes to its various finished products. It then transmits that information to its immediate customers, such as through a dual invoice showing both purchase price and total accumulated emissions in each product sold to a customer.
Read more: Green skills gap: upskilling for climate risk and reporting
This is also where legitimate carbon offsets could have a role, Prof. Kaplan said. “We could subtract any legitimate carbon offset to lower the carbon content of the products a company sells to its immediate customers, which any environmentally sensitive or regulated customer should value,” he said. “By using both our E-liability and E-asset carbon accounting principles, companies around the globe have a strong incentive to invest in legitimate carbon reductions and removal.”
The appeal of the e-liability system is that “it works no matter how complex, distributed or diverse the supply chain. Each entity needs only to know its own emissions and the emissions in the products and services it acquires from its immediate, tier-one suppliers,” Prof. Kaplan said.
“It does the internal allocation process and then transfers, with its invoice, the embedded emissions to its immediate customers. Scope 1 emissions are calculated only once, where they occur, and that’s where they can and should be audited. We can then get clean audits based on observations of the processes that companies are using to measure their acquired and Scope 1 emissions and allocate them to their outputs.”
Another advantage of the system is its primary emphasis on decarbonisation, and not just complying with a disclosure mandate. “Every single company, healthcare institution and university will have guidance for reducing its carbon footprint,” Prof. Kaplan said. “It means manufacturing companies have accountability to redesign their products, re-engineer their processes, and source low-carbon products and services from its suppliers and transportation companies. This is how global decarbonisation will occur.”
Tackling carbon credits
Profs. Kaplan and Ramanna next looked at the challenge of eliminating existing carbon from the atmosphere in the form of e-assets. Producing Scope 1 emissions creates an E-liability: “something we owe back to nature, to restore and reduce the carbon footprint,” he said. “We want to create carbon credits and other e-assets for reducing GHG already in the atmosphere.
According to Prof. Kaplan, application of fundamental accounting principles provides a measurement mechanism to incentivise legitimate carbon capture and sequestration. “Before we addressed this problem, the accounting and measurement in this space was even worse than it was in supply chain carbon accounting,” he recalled. “About 95% of measurements in net zero are fictitious; they’re not actually removing existing carbon from the atmosphere.”
There are two types of offsets, beginning with removal offsets, which Prof. Kaplan described as “real”: “These actually capture existing carbon in the atmosphere and sequester it for an indefinite period of time,” he said, noting that a characteristic of removal offsets is that they required measurement of the duration of carbon capture in addition to their quantity. That’s important in areas like forestry-based capture, as a tree “doesn’t last for the thousand years that CO2 emissions stay in the atmosphere”.
The other type is avoidance offsets, “where organisations promise not to do bad things in the future, get paid for not doing that, and then sell that activity as an offset to somebody who’s currently emitting”, as Prof. Kaplan put it. “We don’t think these are legitimate offsets,” he said.
“One of the principles we’ve developed in our second paper is that only removal offsets can be used to offset an organisation’s acquired and created emissions. Avoidance offsets may be socially desirable to prevent bad things from happening in the future, but preventing something happening in the future is not one of the fundamental concepts in accounting, and therefore it should not be part of an environmental ledger accounting system. We want good removal offsets to be tradable; that enables people with a comparative advantage in doing removal offsets to get funding via capital markets from organisations that truly do want to reduce their environmental footprint.
Read more: Three climate risk challenges (and solutions) for industry
“Organisations creating carbon sequestration have an asset they can sell – that’s what’s being traded, relying on accounting concepts of ‘estimable and probable’ to say when a removal offset is capitalised on an organisation’s carbon balance sheet and eligible for trading to another organisation. And carbon removals that are not estimable or probable cannot yet be considered an asset and should not be part of any voluntary carbon trading system.”
Cost accounting for downstream emissions
Initially, the researchers did not consider downstream emissions at all because of how different they are from the upstream emissions contemplated under Scopes 1 and 2. “A company has, for the most part, no control over what happens to its products or services once it’s sold them to a customer,” according to Prof. Kaplan. However, people kept challenging him and Prof. Ramanna about accountability for downstream emissions, so we “figured there must be something there that we should address”, he said.
“In the last 12 months, we found out what was “there” in downstream emissions, and it was important – not for B2B [business-to-business] companies (because their downstream emissions will be picked up by the e-liability system) but for B2C companies that sell to consumers – because it seems unlikely that we’re going to get 8 billion consumers on our planet to file annual e-ledger statements about their emissions.”
Profs. Kaplan and Ramanna proposed that two conditions should apply for downstream emissions disclosures to be relevant. “Downstream emissions should apply only to B2C [business-to-consumer] companies, and only to B2C companies whose customers actually generate emissions through use of a product,” Prof. Kaplan said. A book, for example, doesn’t generate emissions when a consumer reads it, whereas cooking a purchased package of rice does produce incremental emissions.
“Of course, the real focus of people wanting disclosure of downstream emissions was to get the fossil fuel companies to have accountability for emissions when end-use consumers burn their products to drive their vehicles or heat their homes,” Prof. Kaplan said. But the bigger emissions picture should include automobile and tyre companies, all of which have an influence over the emissions a product generates. “Where’s the accountability for that?”
Again, it’s a problem cost accounting can solve, although in a more speculative way because it depends on disclosure of the rate of carbon per unit of use by the end-use consumer, Prof. Kaplan said.
“It turns out that companies, in their design and sourcing decisions, can control the rate of CO2 emissions per unit of use, but they don’t have control of how many units the consumer uses,” he said. “Nor do we want them dictating to us how much we can use their product.”
Beyond ‘true and fair’
In contrast with the situation when the voluntary GHG Protocol was created two decades ago, the accounting world is “used to universal systems where all organisations operate under that system.” The Australian government’s decision to mandate disclosures is a leading indicator of how the world has begun the migration to a universal carbon accounting system, Prof. Kaplan said. An important next step will be getting to a universal system of genuine environmental accountability.
Profs. Kaplan and Ramanna’s work in promulgating the e-liability system as a means to provide that accountability has included working with global companies, demonstrating the feasibility of their method and assessing its accuracy and auditability. They’ve also been working with enablers such as IT suppliers to develop scalable carbon accounting software, as well as with auditors, as they pilot out how to provide assurance for e-liability reports.
Subscribe to BusinessThink for the latest research, analysis and insights from UNSW Business School
Prof. Kaplan predicted that E-liability and E-asset accounting will be more widely adopted as countries and regions in the global North introduce new tariff regimes based on the carbon content of imported goods. “Then we’ll get a societal good as well as whatever protectionism the country and region desires,” he said.
He also envisions how the GHG protocol can evolve to the e-liability system by having a requirement that a higher percentage of purchases each year be reported using the primary data produced by an e-liability carbon system.
“It’s going to be an iterative, recursive process, and it may take up to five years before everybody is on that system and able to report,” he said. “But you should recognise this transition from a static system that fundamentally is inherently inaccurate and un-auditable, to a true and fair standard, under a carbon accounting system that operates in real-time, with actual carbon being measured and transmitted,” he concluded.