Detecting environmental disclosure management

This paper measures the extent to which firms manipulate their carbon dioxide emission disclosure in corporate social responsibility (CSR) reports. Whereas many studies provide empirical evidence for earnings management around outstanding corporate events such as seasoned equity offerings (Rangan, 1998; Shivakumar, 2000; Teoh et al., 1998) and merger and acquisitions (Erickson and Wang, 1999; Louis, 2004), the quantitative literature is silent with respect to firms’ intention to manage non-financial disclosures from CSR reports. We fill this gap and develop an empirical model to evaluate whether firms manipulate carbon dioxide disclosure in their favor before seasoned equity offerings, after firm-specific environmental disasters, and after being dropped from an internationally renowned list of the Top 100 sustainability firms (“Corporate Knights” list). Our analyses are based on all firms that have carbon dioxide emission data in the Asset4 database from Thomson Reuters. The total sample size is 11,104 observations from 2,263 global firms during the period 2003−2015. We define carbon dioxide emission disclosure management as the residual from a regression of changes in reported direct and indirect carbon dioxide emissions (outputs) on changes on cost of goods sold, changes in depreciation, changes in selling and general administrative expenses and other operating expenses (inputs). We find that audits of sustainability reports do not prevent environmental disclosure management. Thus, external assurance is not able to mitigate the problem of biased environmental disclosures. We attribute this finding to the limited assurance of sustainability report audits. Our findings have implications for the audit profession, which should revise their practice of proving only limited assurance levels on CSR reports and should start the transition from a mere labelling to a thorough auditing process. Our results imply that investors and other stakeholders should consume CSR disclosures with caution and discount their importance due to a profound lack of reliability.