In this article, we will highlight where gaps in corporate carbon emissions data still exist, explore the differences between commonly used estimation models, and share Sustainalytics’ approach to closing the gap in carbon emissions reporting.
Disclosures of scope 1 and 2 emissions, which relate to a company’s direct activities (scope 1) and the first-line indirect emissions associated with purchased electricity (scope 2), are easier to measure. They still make up the majority of reporting on CO2 emissions.
Scope 3 emissions, which cover emissions across a company’s value chain – both upstream in the development and production processes, and downstream in the use phase of a company’s goods and services – are much more difficult to quantify and difficult to disclose . However, based on an analysis of companies in our research universe, the quality and quantity of scope 3 emissions disclosures are increasing faster than those of scope 1 and 2.
Research from Morningstar Sustainalytics shows that for fiscal 2021, nearly 40% of companies in our global ESG universe reported scope 1 and 2 data. This is an increase from the 33% reported in 2020. Scope 3 disclosures for the same universe reached nearly 24% in 2021, up from 19% the year before. Although these figures represent a record high, the downside of these figures is still alarming: 60% of scope 1 and 2 emissions data and over 75% of scope 3 emissions data are not reported.1
The reporting gap is even wider in light of the fact that scope 3 emissions make up more than 60% of total emissions in every industrial sector Sustainalytics covers, with the exception of the utilities sector, where that figure is 57%. In some sectors, such as consumer goods, industry and real estate, scope 3 represents more than 90% of emissions. This underreporting highlights the critical need for a rigorous and scientific estimation model for scope 3 emissions.
The benefits of estimating carbon emissions using a bottom-up approach
Although there are different approaches to estimating CO2 emissions, for the purposes of this article we compare two methods: a bottom-up approach and a top-down approach.
In a bottom-up approach Analysts use the emissions data reported by companies from scope 1, 2 and 3 as the basis for the average of each subsector. They then apply metrics to each company as a measure of size and scale, refining factors that take geography and business model into account.2 This approach forms the basis for Sustainalytics’ enhanced multi-metric, multi-factor (MMMF) regression model to estimate scope 1 and 2 emissions. This model complements our scope 3 estimation model, which also uses a bottom-up approach.
Using company data from the current fiscal year to form a baseline, the models reflect the current economic landscape, taking into account how events such as recessions or pandemics can affect each industry and subsector differently. Using more and more available emissions data, the dynamic nature of this model means that the subsector base averages will be tighter and more accurate each year, resulting in more accurate estimates once the refinement factors are included.
The main alternative to the bottom-up approach is hiring upside downinput-output economic modeling to estimate the business models of different industries.3 Input-output modeling is a well-recognized approach and has an extensive track record in understanding the complex network of interactions between different industrial sectors. For carbon emissions modeling, environmentally friendly input-output (EEIO) models have been developed to account for the associated emissions that arise as a result of the economic interactions between industrial sectors to produce companies’ goods and services.
The disadvantage of EEIO models is that they represent a specific point in time.4 Their results are just a snapshot of the current economic landscape, which is itself dynamic. If the EEIO models are not updated comprehensively and consistently, the accuracy and consistency of the carbon emissions estimation models in which they are used will be diluted. As business models change – technology improves and disrupts, alternatives emerge, consumer behavior changes, etc. – EEIO models less accurately reflect the emissions associated with those business models, thus failing to perform the specific task perform what they were designed for.
Closing the data gap on carbon emissions
While the quality and quantity of CO2 emissions reporting is improving across all three scopes, a significant gap still remains in the disclosure of companies’ greenhouse gas emissions. While several methods exist to help close this gap, a bottom-up MMMF regression model, as used in our Carbon Emissions data, provides a more robust solution to the underreporting of carbon emissions. In an environment where less than a quarter of companies report on scope 1, 2 and 3, it is crucial that investors have accurate, stable and future-proof emissions estimates at their fingertips.
Do you have questions about how your company can address scope 1, 2 and 3 disclosure gaps in your portfolio? Contact our team to find out how Sustainalytics’ Carbon Emission Data solution can help you or speak to your customer advisor.
1 All calculations are based on an analysis of the universe of Sustainalytics’ Carbon Emission Data. 17,407 companies for FY 2020 and 16,725 companies for FY 2021.
2 For example, a country’s electricity generation mix is used as a geographic refinement factor. The carbon intensity of electricity purchased in a country is essentially the scope 2 emissions intensity of all activities in that country.
3 Raynaud, J. 2015. “Carbon Compass: Investor Guide to Carbon Footprinting.” November 23, 2015. Kepler Cheuvreux Transition Research.