Natural Language Processing Analysis of Sustainability Disclosures and Corporate Operation Efficiency

During the past two decades, stakeholders have been pressuring firms to take initiative with their sustainability agendas, leading to a booming demand for firms’ non-financial disclosure. Sustainability performance is increasingly considered to evaluate organizations’ management quality, identify their risk exposure, and assess their ability to leverage business opportunities. Sustainability disclosure not only triggers sustainability actions but also enables a company to showcase its existing sustainability-related activities. A common belief is that if a company is committed to sustainability reporting, its sustainability performance will improve given public exposure and scrutiny. As stakeholders in the capital market reward companies with higher sustainability performance depending on the quality of disclosures, a relationship between sustainability data quality and corporate performance arises. Amid the confusing plethora of reporting guidelines, companies have long struggled to produce reliable data on sustainability performance. To investigate the data value of sustainability reports, this study takes stock of the current status of sustainability reporting using natural language processing methodology. After identifying prominent characteristics of the sustainability data reported, the study investigates and finds a positive relationship between these characteristics and corporate operational efficiency proxied by firm-year level and industry-controlled efficiency scores computed using data envelopment analysis. Further analysis shows that this positive relationship is more pronounced for firms with better external monitoring mechanisms in terms of social, environment, and governance ratings than other firms. This study concludes that sustainability reporting positively affects corporate operational efficiency and external monitoring mechanisms matter.