Most corporations are not reporting the full scope of their carbon footprint with many claiming to be ‘green’ despite a lack of reporting on Scope 3 key categories, a new study shows.
Research fellow Dr Quyen Nguyen, of the University of Otago’s Climate and Energy Finance Group, conducted the research in collaboration with Griffith University and climate risk analysis firm EMMI.
Though CO2 reporting is voluntary for most firms, corporations are under pressure from investors,regulators, politicians, non-profit organisations and other stakeholders to disclose and reduce greenhouse gas emissions (GHG).
A statement from the researchers explains the standard for greenhouse gas accounting, the Greenhouse Gas Protocol, is used worldwide to measure a company’s total carbon footprint with three levels of reporting.
- The first measures the GHG emissions directly produced by a company during business activities (such as emissions from a corporate fleet).
- The second measures emissions associated with the production of energy which is purchased from an external supplier (such as emissions produced by electricity providers).
- The third (Scope 3) measures indirect emissions not already accounted for and includes upstream and downstream emissions from a company’s full value chain, such as emissions produced by customers as a result of a company’s product (downstream) and emissions produced in the manufacture of a company’s equipment (upstream).
Nguyen says the researchers used machine learning to improve the prediction of corporate carbon footprints, which provided an indication of where policymakers and regulators should concentrate their efforts for greater disclosure.
“We discovered firms chose to report on certain categories within Scope 3 and they often chose to report on categories which are easier to calculate instead of categories which really matter like use of sold products.
“Firms generally report incomplete compositions of Scope 3 emissions, yet they are reporting more categories over time.
“It is interesting to see the Scope 3 categories firms choose to report on are not always the most material, such as travel emissions and this may be because it is difficult to collect data for other relevant and material categories (such as the use of products and processing of sold products), but it could also mean that the true environmental impact of a firm is being disguised.”
Nguyen says machine learning can help predict individual Scope 3 categories, “but it is no magic bullet, what we need is for firms to report more Scope 3 categories”.
“Firms are reporting more categories over time, and the fraction of firms which report scope 3 emissions are around 60 percent of firms which are already reporting Scope One and Two emissions.”
Co-author Professor Ivan Diaz-Rainey, an international expert in climate and sustainable finance from Griffith’s Department of Accounting, Finance and Economics and previously the University of Otago, says firms were being strategic in their Scope 3 reporting and this could underpin greenwashing.
“Scope 3 emissions account for the highest proportion of total emissions, and it’s the least likely scope to be reported on,” he says.
“Companies have a great incentive to better their scope one and two emissions because direct energy efficiency leads to financial savings.
Diaz-Rainey says some jurisdictions are moving towards mandatory disclosures, driven by the Task Force on Climate-Related Financial Disclosures (TCFD), and pressure to make Scope 3 mandatory is increasing.
UNSW Climate Change Research Centre adjunct fellow and co-founder of EMMI Dr Ben McNeil says Scope 3 emissions for companies were difficult to quantify but critical in understanding how companies were financially exposed to carbon pricing and their decarbonisation pathways.
“Although significant uncertainty remains, our novel machine learning approach to estimating Scope 3 emissions has proven valuable to understand whether a company has ‘material’ financial exposure to a net-zero world where carbon is legislated and priced,” McNeil says.
Publication details:
Scope 3 emissions: Data quality and machine learning prediction accuracy