Metadata integrity in digital disclosures: a multilevel study of financial and ESG reporting structures
摘要
This study examines the structural, organizational, and contextual determinants of metadata integrity in digital corporate disclosures, covering both financial and ESG reporting. It evaluates how reporting format, ESG inclusion, firm size, sector, and national regulatory environments shape metadata incoherence—operationalized through a Metadata Incoherence Score (MIS) that captures deficiencies in completeness, consistency, and semantic coherence across full reports. Using 700 disclosures from five countries and a multilevel Generalized Linear Mixed Model, the analysis shows that machine‑readable formats (HTML and especially XHTML/XBRL) significantly reduce incoherence relative to PDF, whereas ESG content increases semantic variability. Firm size and sector effects remain significant, and cross‑country differences persist, highlighting institutional variation in digital reporting governance. The study offers a replicable framework for diagnosing metadata integrity and supports a staged policy pathway combining mandatory baseline tagging with best‑practice ESG templates and validation tools to improve digital reporting quality.