Human Trafficking in the Global Supply Chain: Using Machine Learning to Understand Corporate Disclosures Under the UK Modern Slavery Act

• Published in: Harvard human rights journal 2022-04, Vol.35 (1), p.1
• Main Authors: Nersessian, David, Pachamanova, Dessislava
• Format: Article
• Language: English
• Subjects: Companies; Human rights; Human trafficking; International law; Machine learning; Public opinion; Slavery; Supply chains
• ISSN: 1057-5057; 1943-5088

This Article applies machine learning techniques to international human rights law and corporate reporting. Specifically, it proposes a new methodological framework to analyze how businesses interpret statutory compliance and respond to statutory ambiguity in the context of mandatory corporate disclosures on human trafficking in the global supply chain. The Article focuses on the United Kingdom's Modern Slavery Act ("MSA") 2015, which requires companies to disclose publicly the risks of human trafficking in their supply chains. While the MSA does not specify clear reporting standards, it seeks to reduce trafficking by creating accountability in the court of "public opinion" through disclosure. The Article tests the utility of this approach by applying advanced machine learning techniques (i.e., natural language processing) to a proprietary dataset created by the Authors of over 17,000 MSA statements released by companies between 2016 and 2019. The analysis reveals that compliance with the statutory mandate is limited and that other aspects of the statute are not uniformly interpreted. Many companies also "anchor" their statements in broader human rights language and emphasize engagement by senior company leaders in an effort to bolster their social reputations. More rigorous enforcement and clearer reporting standards thus are needed to strengthen the MSA and similar reporting laws in other jurisdictions in order to achieve the statutory purpose of redressing human trafficking in the global supply chain.