Machine Learning Models for Prediction of Scope 3 Carbon Emissions

36 Pages Posted: 20 Jul 2022

See all articles by George Serafeim

George Serafeim

Harvard Business School

Gladys Velez Caicedo

Harvard Business School

Date Written: June 1, 2022

Abstract

For most organizations, the vast amount of carbon emissions occur in their supply chain and in the post-sale processing, usage, and end of life treatment of a product, collectively labelled scope 3 emissions. In this paper, we train machine learning algorithms on 15 reported types of scope 3 emissions. The models utilize as inputs widely available financial statement variables, scope 1 and 2 emissions, and industrial classifications. We find that most reported scope 3 emission types can be predicted with higher accuracy using Adaptive Boosting machine learning algorithms relative to linear regression models and other supervised machine learning algorithms.

Suggested Citation

Serafeim, George and Velez Caicedo, Gladys, Machine Learning Models for Prediction of Scope 3 Carbon Emissions (June 1, 2022). Harvard Business School Accounting & Management Unit Working Paper No. 22-080, Available at SSRN: https://ssrn.com/abstract=4149874 or http://dx.doi.org/10.2139/ssrn.4149874

George Serafeim (Contact Author)

Harvard Business School ( email )

Boston, MA 02163
United States

HOME PAGE: http://www.hbs.edu/faculty/Pages/profile.aspx?facId=15705

Gladys Velez Caicedo

Harvard Business School ( email )

Boston, MA 02163
United States

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
887
Abstract Views
2,045
rank
39,999
PlumX Metrics